1 This is aspell.info, produced by makeinfo version 4.8 from aspell.texi.
3 This is the user's manual for Aspell
5 GNU Aspell is a spell checker designed to eventually replace Ispell.
6 It can either be used as a library or as an independent spell checker.
8 Copyright (C) 2000-2011 Kevin Atkinson.
10 Permission is granted to copy, distribute and/or modify this
11 document under the terms of the GNU Free Documentation License,
12 Version 1.1 or any later version published by the Free Software
13 Foundation; with no Invariant Sections, no Front-Cover Texts and
14 no Back-Cover Texts. A copy of the license is included in the
15 section entitled "GNU Free Documentation License".
17 INFO-DIR-SECTION GNU Packages
19 * Aspell: (aspell). GNU Aspell spelling checker
23 File: aspell.info, Node: Top, Next: Introduction, Up: (dir)
28 This is the user's manual for Aspell
30 GNU Aspell is a spell checker designed to eventually replace Ispell.
31 It can either be used as a library or as an independent spell checker.
38 * Customizing Aspell::
39 * Working With Dictionaries::
40 * Writing programs to use Aspell::
41 * Adding Support For Other Languages::
42 * Implementation Notes::
43 * Languages Which Aspell can Support::
44 * Language Related Issues::
51 --- The Detailed Node Listing ---
55 * Spellchecking Individual Files::
56 * Using Aspell as a Replacement for Ispell::
57 * Using Aspell with other Applications::
61 * Specifying Options::
63 * Dumping Configuration Values::
64 * Notes on Various Options::
66 Notes on Various Options
68 * Notes on Various Filters and Filter Modes::
69 * Notes on the Prefix Option::
70 * Notes on Typo-Analysis::
71 * Notes on the Different Suggestion Modes::
73 Working With Dictionaries
75 * Using aspell-import::
76 * How Aspell Selects an Appropriate Dictionary::
77 * Listing Available Dictionaries::
78 * Dumping the Contents of the Word List::
79 * Creating an Individual Word List::
80 * Working With Affix Info in Word Lists::
81 * Format of the Personal and Replacement Dictionaries::
82 * Using Multi Dictionaries::
86 Writing programs to use Aspell
90 * Notes on Storing Replacement Pairs::
92 Adding Support For Other Languages
94 * The Language Data File::
95 * Compiling the Word List::
97 * The Simple Soundslike::
98 * Replacement Tables::
100 * Controlling the Behavior of Run-together Words::
101 * Creating A New Character Set::
102 * Creating An Official Dictionary Package::
106 * Aspell Suggestion Strategy::
107 * Notes on 8-bit Characters::
109 Languages Which Aspell can Support
114 * Planned Dictionaries::
117 Language Related Issues
120 * Words With Symbols in Them::
121 * Unicode Normalization::
123 * Context Sensitive Spelling::
129 * Notes on Various Items::
131 Notes on Various Items
133 * Word skipping by context::
134 * Hidden Markov Model::
135 * Email the Personal Dictionary::
139 * Generic Install Instructions::
140 * HTML Manuals and "make clean"::
142 * Loadable Filter Notes::
143 * Upgrading from Aspell 0.50::
144 * Upgrading from Aspell .33/Pspell .12::
145 * Upgrading from a Pre-0.50 snapshot::
150 * GNU Free Documentation License::
151 * GNU Lesser General Public License::
154 File: aspell.info, Node: Introduction, Next: Support, Prev: Top, Up: Top
159 GNU Aspell is a spell checker designed to eventually replace Ispell. It
160 can either be used as a library or as an independent spell checker. Its
161 main feature is that it does a much better job of suggesting possible
162 replacements for a misspelled word than just about any other spell
163 checker out there for the English language. Unlike Ispell, Aspell can
164 also easily check documents in UTF-8 without having to use a special
165 dictionary. Aspell will also do its best to respect the current locale
166 setting. Other advantages over Ispell include support for using
167 multiple dictionaries at once and intelligently handling personal
168 dictionaries when more than one Aspell process is open at once.
170 The latest version of Aspell can always be found at
173 1.1 Comparison to other spell checker engines
174 =============================================
176 Aspell Ispell Netscape Microsoft
179 Suggestion 88-98 54 55-70? 71
183 Alternate Dictionaries x x ? ?
184 International Support x x ? ?
186 The Suggestion Intelligence is based on a small test kernel of
187 misspelled/correct word pairs. Go to `http://aspell.net/test' for more
188 info and how you can help contribute to the test kernel. The current
189 scores for Aspell are 88 in _fast_ mode, 93 in _normal_ mode, and 98 in
190 _bad spellers_ mode: for more information about the various suggestion
191 modes *Note Notes on the Different Suggestion Modes::.
193 If you have any other information you would like to add to this chart
194 please contact me at <kevina@gnu.org>.
196 1.1.1 Comparison to Ispell
197 --------------------------
199 1.1.1.1 Features that only Aspell has
200 .....................................
202 * Is an actual library that other programs can link to instead of
203 having to use it through a pipe.
205 * Does a much better job of suggesting possible replacements for a
206 misspelled word than Ispell does or for that matter many other
207 spell checkers I have seen. If you know a spell checker that does
208 a better job please let me know.
210 * Can learn from user's misspellings.
212 * Can easily check documents in UTF-8 without having to use a special
215 * Has support for using multiple dictionaries at once.
217 * Is multiprocess intelligent. When a personal dictionary (or
218 replacement list) is saved, it will now first update the list
219 against the dictionary on disk in case another process modified it.
221 * Can share the memory used in the main word list between processes.
223 * A better, more complete word list for the English language. Word
224 lists are provided for American, British, and Canadian spelling.
225 Special care has been taken to only include one spelling for each
226 word in any particular word list. The word list included in
227 Ispell by contrast only included support for American and British
228 and also tends to included multiple spellings for a word which can
229 mask some spelling errors.
231 1.1.1.2 Things that, currently, only Ispell has
232 ...............................................
234 * Lower memory footprint
236 * Ability to deal with arbitrary multi-character letters such as old
237 ASCII encodings of accented letters.
239 * Perhaps better support for spell checking (La)TeX files.
242 For a detailed description of how Aspell differs from Ispell, *Note
243 Differences From Ispell::.
246 File: aspell.info, Node: Support, Next: Basic Usage, Prev: Introduction, Up: Top
251 Support for Aspell can be found on the Aspell mailing lists.
252 Instructions for joining the various mailing lists (and an archive of
253 them) can be found off the Aspell home page at `http://aspell.net'.
255 Bug reports should be submitted via the Sourceforge Tracker at
256 `http://sourceforge.net/tracker/?group_id=245' rather than being posted
257 to the mailing lists.
260 File: aspell.info, Node: Basic Usage, Next: Customizing Aspell, Prev: Support, Up: Top
265 For a quick reference on the Aspell utility use the command `aspell
270 * Spellchecking Individual Files::
271 * Using Aspell as a Replacement for Ispell::
272 * Using Aspell with other Applications::
275 File: aspell.info, Node: Spellchecking Individual Files, Next: Using Aspell as a Replacement for Ispell, Up: Basic Usage
277 3.1 Spellchecking Individual Files
278 ==================================
280 To use Aspell to spellcheck a file use:
282 aspell check [OPTIONS] FILENAME
284 at the command line where `FILENAME' is the file you want to check and
285 `OPTIONS' is any number of optional options. Some of the more useful
289 the mode to use when checking files. The available modes are
290 `none', `url', `email', `sgml', `tex', `texinfo', `nroff', among
291 others. For more information on the various modes see *Note Notes
292 on Various Filters and Filter Modes::.
295 don't create a backup file. Normally, if there are any corrections
296 the Aspell utility will append `.bak' to the existing file name
297 and then create a new file with corrections made during spell
301 the suggestion mode to use where mode is one of `ultra', `fast',
302 `normal', or `bad-spellers'. For more information on these modes
303 see *Note Notes on the Different Suggestion Modes::.
306 the language the document is written in. The default depends on
310 encoding the document is expected to be in. The default depends
311 on the current locale.
314 the main dictionary to use.
317 the keymapping to use. Either `aspell' for the default mapping or
318 `ispell' to use the same mapping that the Ispell utility uses.
320 For more information on the available options, please see *Note
321 Customizing Aspell::.
323 For example to check the file `foo.txt':
327 and to check the file `foo.txt' using the `bad-spellers' suggestion
328 mode and the American English dictionary:
330 aspell check --sug-mode=bad-spellers -d en_US foo.txt
332 If the `mode' option is not given, then Aspell will use the
333 extension of the file to determine the current mode. If the extension
334 is `.tex', then `tex' mode will be used, if the extension is `.html',
335 `.htm', `.php', or `.sgml' it will check the file in `sgml' mode,
336 otherwise it will use `url' mode.
338 For more information on the various modes that can be used, see
339 *Note Notes on Various Filters and Filter Modes::.
341 If Aspell was compiled with curses support and the `TERM'
342 environment variable is set to a capable terminal type then Aspell will
343 use a nice full screen interface, otherwise it will use a simpler
344 "dumb" terminal interface where the misspelled word is surrounded by
345 two '*'. In either case the interface should be self explanatory.
347 If Aspell is compiled with a version of the curses library that
348 support wide characters then Aspell can also check UTF-8 text.
349 Furthermore, the document will be displayed in the encoding defined by
350 the current locale. This encoding does not necessarily have to be the
351 same encoding that the document is in. This means that is is possible
352 to check an 8-bit encoding such as ISO-8859-1 on an UTF-8 terminal. To
353 do so simply set the `encoding' option to `iso-8859-1'. Furthermore it
354 is also possible to check an UTF-8 document on an 8-bit terminal
355 provided that the document can be successfully converted into that
359 File: aspell.info, Node: Using Aspell as a Replacement for Ispell, Next: Using Aspell with other Applications, Prev: Spellchecking Individual Files, Up: Basic Usage
361 3.2 Using Aspell as a Replacement for Ispell
362 ============================================
364 As of GNU Aspell 0.60.1 Aspell should be able to completely replace
365 Ispell for most applications. The Ispell compatibility script should
366 work for most applications which expect Ispell. However there are some
367 differences which you should be aware of.
369 3.2.1 As a Drop In Replacement
370 ------------------------------
372 Aspell can be used as a drop in replacement for Ispell for programs
373 that use Ispell through a pipe such as Emacs and LyX. It can also be
374 used with programs that simply call the `ispell' command and expect the
375 original file to be overwritten with the corrected version.
377 If you do not have Ispell installed on your system and have installed
378 the Ispell compatibility script then you should not need to do anything
379 as most applications that expect Ispell will work as expected with
380 Aspell via the Ispell compatibility script.
382 Otherwise, the recommended way to use Aspell as a replacement for
383 Ispell is to change the `ispell' command from within the program being
384 used. If the program uses `ispell' in pipe mode simply change `ispell'
385 to `aspell'. If the program calls the `ispell' command to check the
386 file, then change `ispell' with `aspell check'.
388 If that is impossible then the `run-with-aspell' script can be used.
389 This script modifies the patch so that programs see the Ispell
390 compatibility script instead of the actual true `ispell' command. The
391 format of the script is:
393 run-with-aspell COMMAND
395 where COMMAND is the name of the program with any optional arguments.
397 The old method of mapping Ispell to Aspell is discouraged because it
398 can create compatibility problems with programs that actually require
399 Ispell such as Ispell's own scripts.
401 3.2.2 Differences From Ispell
402 -----------------------------
404 Nevertheless, Aspell is not Ispell, nor is it meant to completely
405 emulate the behavior of Ispell. The `aspell' command is not identical
406 to the `ispell' command when not used in "pipe" mode. If an
407 application expects the `ispell' command, then the Ispell compatibility
408 script should be used instead.
410 3.2.2.1 Functionality of the Ispell Compatibility Script
411 ........................................................
413 The Ispell compatibility script provides the following Ispell
416 * The ability to check a file when called without any mode
419 * The pipe or -a mode.
421 * The list or -l mode.
423 * The version or -v mode. A single line is returned which, while not
424 being identical to the line Ispell returns, is sufficient to fool
427 * The munch or -c mode.
429 * The expand or -e mode.
431 * The ability to dump the affix file when called with '-D'. However
432 the format of the affix file is different. Furthermore, not all
433 languages have an affix file.
436 However the Ispell script is currently unable to emulate the '-A'
437 pipe mode. This is different from the normal pipe mode in that the
438 special `&Include_File&' command is recognized.
440 3.2.2.2 Recognized Options
441 ..........................
443 Aspell, and thus the Ispell compatibility script, recognizes most of
444 the options that Ispell uses except for the '-S', '-w' and '-T'
445 options. The Aspell command will simply ignore these options if it
448 3.2.2.3 Check Mode Compatibility
449 ................................
451 The interface used by Aspell when checking individual files is slightly
452 different than Ispell's. In particular the default keymappings are not
453 the same as the ones Ispell uses. However Aspell supports using the
454 Ispell keymappings via the `keymapping' option. To use the Ispell
455 keymappings set the `kepmapping' option to `ispell'. This can be done
456 on the command line by adding using the command:
458 aspell check --keymapping=ispell ...
459 or with the Ispell compatibility script
460 ispell --keymapping=ispell ...
462 The Ispell keymapping can always be used when the Ispell compatibly
463 script is called by uncommenting the indicated line in the `ispell'
466 3.2.2.4 Pipe Mode Compatibility
467 ...............................
469 The Aspell pipe mode should be identical to the Ispell pipe mode except
470 if the line starts with a '$$' as that will trigger special Aspell only
471 commands or if the line starts with a '~' which is ignored by Aspell.
473 3.2.2.5 Other Differences
474 .........................
476 The compiled dictionary format is completely different than Ispell's.
477 Furthermore the format of the language data files' are different than
478 Ispell's affix file. However, all known Ispell dictionaries were
479 converted to Aspell format, except for Albanian (sq) as I was unable to
480 find the source word list.
482 The naming and format of the personal dictionary is also different.
483 However, Ispell personal dictionaries can be imported using the
484 `aspell-import' script. *Note Using aspell-import::. The Ispell
485 personal dictionary is simply a list of words while the Aspell one is a
486 list of words with a header line. Thus it is also fairly easy to
487 convert between the two. *Note Format of the Personal Dictionary::.
489 3.2.2.6 Missing Functionally
490 ............................
492 The only major area where Ispell is superior to Aspell is in the
493 handling of multi character letters such as old ASCII encoding of
496 However, Aspell can handle UTF-8 documents far better than Ispell
500 File: aspell.info, Node: Using Aspell with other Applications, Prev: Using Aspell as a Replacement for Ispell, Up: Basic Usage
502 3.3 Using Aspell with other Applications
503 ========================================
505 3.3.1 With Emacs and XEmacs
506 ---------------------------
508 The easiest way to use Aspell with Emacs or XEmacs is to add this line:
511 (setq-default ispell-program-name "aspell")
513 to the end of your `.emacs' file.
515 For some reason version 3.0 of ispell.el (the lisp program that
516 (x)emacs uses) wants to reverse the suggestion list. To fix this add
520 (setq-default ispell-extra-args '("--reverse"))
522 after the previous line in your .emacs file and it should solve the
525 Ispell.el, version 3.1 (December 1, 1998) and better, has the list
526 reversing problem fixed. You can find it at
527 `http://www.kdstevens.com/~stevens/ispell-page.html'.
532 Version 1.0 of LyX provides support for Aspell's learning from user's
535 To use Aspell with LyX 1.0 either change the `spell_command' option
536 in the `.lyxrc' file or use the `run-with-aspell' utility.
541 To use Aspell in VIM you simply need to add the following line to your
545 map ^T :w!<CR>:!aspell check %<CR>:e! %<CR>
547 I use `Ctrl-T' since that's the way you spell check in `pico'. In
548 order to add a control character to your `.vimrc' you must type
549 `Ctrl-v' first. In this case `Ctrl-v Ctrl-t'.
551 A more useful way to use Aspell, IMHO, is in combination with
552 Newsbody (`http://www.image.dk/~byrial/newsbody/') which is how I use it
553 since VIM is my editor for my mailer and my news reader.
555 map ^T\\1\\2<CR>:e! %<CR>
557 map \\2 :!newsbody -qs -n % -p aspell check \\%f<CR>
562 To use Aspell in Pine simply change the option `speller' to
564 aspell --mode=email check
566 To change the `speller' option go to the main menu. Type `S' for
567 _setup_, `C' for _config_, then `W' for _where is_. Type in `speller'
568 as the word to find. The speller option should be highlighted now.
569 Hit enter, type in the above line, and hit enter again. Then type `E'
570 for _exit setup_ and `Y' to save the change.
572 If you have a strong desire to check other people's comments change
577 instead which will avoid switching Aspell into email mode.
580 File: aspell.info, Node: Customizing Aspell, Next: Working With Dictionaries, Prev: Basic Usage, Up: Top
585 The behavior of Aspell can be changed by any number of options which
586 can be specified at either the command line, the environment variable
587 `ASPELL_CONF', a personal configuration file, or a global configuration
588 file. Options specified on the command line override options specified
589 by the environment variable. Options specified by the environment
590 variable override options specified by either of the configuration
591 files. Finally options specified by the personal configuration file
592 override options specified in the global configuration file. Options
593 specified in the environment variable `ASPELL_CONF', a personal
594 configuration file, or a global configuration file will take effect no
595 matter how Aspell is used which includes being used by other
598 Aspell has three basic types of options: "boolean", "value", and
601 "Boolean" options are either enabled or disabled, "value" options
602 take a specific value, and "list" options can either have entries added
603 or removed from the list.
607 * Specifying Options::
609 * Dumping Configuration Values::
610 * Notes on Various Options::
613 File: aspell.info, Node: Specifying Options, Next: The Options, Up: Customizing Aspell
615 4.1 Specifying Options
616 ======================
618 4.1.1 At the Command Line
619 -------------------------
621 All options specified at the command line have the following basic
626 where the `=' can be replaced by whitespace.
628 Some options also have single letter abbreviations of the form:
630 LETTER [OPTIONAL_WHITESPACE VALUE]
632 Any non-ASCII characters are expected to be in the encoding
633 specified by the current locale.
635 To reset an option to the default value, prefix the option with a
636 `reset-' and don't specify a value.
638 4.1.1.1 Value options
639 .....................
641 To specify a value option simply specify the option with its
642 corresponding value. For example to set the filter mode to TeX use
645 If a value option has a single letter shortcut simply specify the
646 single letter shortcut with its corresponding value. For example to
647 use a the accented version of the American English dictionary use `-d
650 4.1.1.2 Boolean options
651 .......................
653 To enable a boolean option simply specify the option without any
654 corresponding value, or prefix it with an `enable-'. For example to
655 create a backup file use `--backup'. To disable a boolean option
656 prefix the option name with a `dont-' or `disable-'. To avoid creating
657 a backup file use `--dont-backup'. Boolean options can also be set
658 directly like a value option where the value is either "true" or
659 "false", for example `--backup=true'.
661 If a boolean option has a single letter abbreviation simply give the
662 letter corresponding to either enabling or disabling the option without
663 any corresponding value. For example, to consider run-together words
664 valid use `-C' or to consider them invalid use `-B'
669 To add a value to the list, prefix the option name with an `add-' and
670 then specify the value to add. For example, to add the URL filter use
671 `--add-filter url'. To remove a value from a list option, prefix the
672 option name with a `rem-' and then specify the value to remove. For
673 example, to remove the URL filter use `--rem-filter url'. To remove
674 all items from a list prefix the option name with a `clear-' without
675 specify any value. For example, to remove all filters use
678 A list option can also be set directly, in which case it will be set
679 to a single value. To directly set a list option to multiple values
680 prefix the option name with a `lset-' and separate each value with a
681 `:'. For example, to use the URL and TeX filter use `--lset-filter
684 4.1.2 Via a Configuration File
685 ------------------------------
687 Aspell can also accept options via a personal or global configuration
688 file. The exact files to used are specified by the options `per-conf'
689 and `conf' respectively but the personal configuration file is normally
690 `.aspell.conf' located in the `HOME' directory and the global one is
691 normally `aspell.conf' which is located in the `etc' directory which is
692 normally `/usr/etc' or `/usr/local/etc'. To find out the particular
693 values for your particular system use `aspell dump config'.
695 Each line of the configuration file has the format:
699 There may be any number of spaces between the option and the value
700 however it can only be spaces, i.e. there is no `=' between the option
701 name and the value and there are no preceding `--' as used on the
704 Comments may also be included by preceding them with a `#' as
705 anything from a `#' to a newline is ignored. Blank lines are also
708 To include a literal `#' use `\#'. To include a literal `\' use
709 `\\'. Any other non-alpha character can also be protected by a `\' if
712 Any non-ASCII characters are expected to be in UTF-8.
714 To reset an option to the default value prefix the option with a
715 `reset-' and don't specify a value.
717 Values set in the personal configuration file override those in the
718 global file. Options specified at either the command line or via an
719 environment variable override those specified by either configuration
722 Note: Filters and corresponding options also may be assembled
723 inside a special meta filter file named `METAFILTER.flt'. A
724 filter has to be loaded via adding a `add-filter FILTERNAME' line
725 to the meta filter file before its options may be specified.
727 4.1.2.1 Value options
728 .....................
730 To specify a value option simply include the option followed by the
731 corresponding value. For example to set the default language to German
734 4.1.2.2 Boolean options
735 .......................
737 To specify a boolean option simply include the option followed by a
738 `true' to enable it or a `false' to disable it. For example to allow
739 run-together words use `run-together true'.
744 To add a value to the list, prefix the option name with an `add-' and
745 then specify the value to add. For example to add the URL filter use
746 `add-filter url'. To remove a value from a list option prefix the
747 option name with a `rem-' and then specify the value to remove. For
748 example, to remove the URL filter use `rem-filter url'. To remove all
749 items from a list prefix the option name with a `clear-' without
750 specifying any value. For example, to remove all filters use
753 A list option can also be set directly, in which case it will be set
754 to a single value. To directly set a list option to multiple values
755 prefix the option name with a `lset-' and separate each value with a
756 `:'. For example, to use the URL and TeX filter use `lset-filter
757 url:tex'. To include a literal `:' use `\:'.
759 4.1.3 Setting Options via an Environment Variable
760 -------------------------------------------------
762 The environment variable `ASPELL_CONF' may also be used and it
763 overrides any options set in the configuration file. The format of the
764 string is exactly the same as the configuration file except that
765 semicolons (`;') are used instead of newlines.
768 File: aspell.info, Node: The Options, Next: Dumping Configuration Values, Prev: Specifying Options, Up: Customizing Aspell
773 The following is a list of available options broken down by category.
774 Each entry has the following format:
776 OPTION[,SINGLE-LETTER-ABBREVIATION]
779 Where single letter options are specified as they would appear at the
780 command line, ie with the preceding dash. Boolean single letter
781 options are specified in the following format:
783 -<abbreviation to enable>|-<abbreviation to disable>
785 OPTION is one of the following: _boolean_, _string_, _file_, _dir_,
786 _integer_, or _list_.
788 _String_, _file_, _dir_, and _integer_ types are all value options
789 which can only take a specific type of value.
791 4.2.1 Dictionary Options
792 ------------------------
794 The following options may be used to control which dictionaries to use
795 and how they behave (for more information see *Note How Aspell Selects
796 an Appropriate Dictionary::):
799 (string) Base name of the dictionary to use. If this option is
800 specified then Aspell will either use this dictionary or die.
803 (dir) Location of the main word list.
806 (string) Language to use. It follows the same format of the `LANG'
807 environment variable on most systems. It consists of the two
808 letter ISO 639 language code and an optional two letter ISO 3166
809 country code after a dash or underscore. The default value is
810 based on the value of the `LC_MESSAGES' locale.
813 (string) The preferred size of the word list. This consists of a
814 two char digit code describing the size of the list, with typical
815 values of: 10=tiny, 20=really small, 30=small, 40=med-small,
816 50=med, 60=med-large, 70=large, 80=huge, 90=insane.
819 (list) Any extra information to distinguish two different words
820 lists that have the same lang and size.
823 (list) Search path for word list information files.
826 (file) Personal word list file name.
829 (file) Replacements list file name.
832 (list) Extra dictionaries to use.
835 (list) create dictionary aliases. Each entry has the form `FROM
836 TO'. Will override any system dictionaries that are present.
839 4.2.2 Encoding Options
840 ----------------------
842 These options control the encoding the document is expected to be in and
846 (string) The encoding the input text is in. Valid values include,
847 but not limited to, `iso-8859-*', `utf-8', `ucs-2', `ucs-4'. When
848 using the Aspell utility the default encoding is based on the
849 current locale. Thus if your locale currently uses the `utf-8'
850 encoding than everything will be in UTF-8. The `ucs-2' and
851 `ucs-4' encodings are intended to be used by other programs using
852 the Aspell library and is not supported by the Aspell utility.
855 (boolean) Perform Unicode normalization. Enabled by default.
858 (boolean) Avoid lossy conversions when normalizing. Lossy
859 conversions includes compatibility mappings such as splitting the
860 letter `OE' (U+152) into `O' and `E' (when the combined letter is
861 not available), and mappings which will remove accents. Disabled
862 by default except when creating dictionaries.
865 (string) The normalization form the output should be in. This
866 option primarily effects the normalization form of the suggestions
867 as when spell checkering as the actual text is unchanged unless
868 there is an error. Valid values are `none', `nfd' for fully
869 decomposition (Normalization Form D), `nfc' for Normalization Form
870 C, or `comp' for fully composed. `comp' is like `nfc' except that
871 _full_ composition is used rather than _canonical_ composition.
872 The `normalize' option must be enabled for this option to be used.
875 (boolean) Set to true when the current language requires Unicode
876 normalization. This is generally the case when private use
877 characters are used internally by Aspell or when Normalization
878 Form C is not the same as full composition.
881 4.2.3 Checker Options
882 ---------------------
884 These options control the behavior of Aspell when checking documents.
887 (integer) Ignore words with N characters or less
890 (boolean) Ignore commands to store replacement pairs.
893 (boolean) Save the replacement word list on save all.
896 (file) The base name of the keyboard definition file to use (*note
897 Notes on Typo-Analysis::)
900 (mode) Suggestion mode = `ultra' | `fast' | `normal' | `slow' |
901 `bad-spellers' (*note Notes on the Different Suggestion Modes::)
904 (boolean) Ignore case when checking words.
907 (boolean) Ignore accents when checking words - _currently ignored_.
913 These options modify the behavior of the Aspell filter interface in
914 general (for more information see *note Notes on Various Filters and
918 (list) filters to use
921 (list) Where to look when loading filter and filter modes.
924 (string) Sets the filter mode. Possible values include, but not
925 limited to, `none', `url', `email', `sgml', or `tex'. (The
926 shortcut options `-e' may be used for email, `-H' for HTML, or
930 These options belong to filters packaged along with Aspell standard
931 distribution. These options may be prefixed by the keyword `f-' in
932 order to explicitly indicate that they are options recognized by a
933 filter and not by Aspell itself.
938 This filter hides quoting characters and email preamble and other parts
939 of an email which need not to be spell checked.
942 (list) Email quote characters.
945 (integer) The number of characters that can appear before the
951 This filter converts an HTML source file into a format which eases
952 spell checking of HTML texts by Aspell.
955 (list) HTML attributes to always check, such as alt= (alternate
959 (list) HTML tags to always skip the contents of, such as <script>.
964 This filter is identical to the HTML filter except that its options has
965 different default values which are currently the empty list.
970 This filter hides all LaTeX commands and corresponding parameters not
971 being readable text in LaTeX output from Aspell.
977 (boolean) check TeX comments
983 This filter hides all Texinfo commands from Aspell. It can also hide
984 Texinfo parameters and environments not corresponding to readable text.
987 (list) Texinfo command to ignore the parameters of.
990 (list) Texinfo environments to ignore.
996 This filter can be used to spell check source codes, HTML sources and
997 other texts which consist of different contexts. These contexts must
998 be separated by pairs of unique delimiters. The different contexts may
999 not be dependent upon each other except for initial context which is
1000 assumed if not any other context applies.
1002 context-visible-first
1003 (boolean) Switches the context which should be visible to Aspell.
1004 Per default the initial context is assumed to be invisible as one
1005 would expect when spell checking source files of programs where
1006 relevant parts are contained in string constants and comments but
1007 not in the remaining code. If set to true the initial context is
1008 visible while the delimited ones are hidden.
1010 add|rem-context-delimiters
1011 (list) Add or remove pairs of delimiters. This allows you to
1012 specify the character, or sequences of characters, which should be
1013 used to switch contexts and therefore have to be escaped by `\' if
1014 they should appear literally. The two delimiting chars belonging
1015 to one pair have to be separated by a space character. If
1016 multiple pairs are specified by one `add|rem-context-delimiters'
1017 call the different pairs have to be separated by a literal comma.
1018 Per default the delimiters are set to C/C++ comment and string
1019 constant delimiters. If the end of line delimits a context than
1020 this has to be indicated by the literal `\0' string.
1022 4.2.5 Run-together Word Options
1023 -------------------------------
1025 These may be used to control the behavior of run-together words (for
1026 more information *note Controlling the Behavior of Run-together
1029 (boolean) consider run-together words valid
1032 (integer) maximum number of words that can be strung together
1035 (integer) minimal length of interior words
1037 4.2.6 Miscellaneous Options
1038 ---------------------------
1040 Miscellaneous other options that don't fall under any other category
1042 (file) Main configuration file. This file overrides Aspell's
1046 (dir) location of main configuration file
1049 (dir) location of language data files
1052 (dir) alternative location of language data files. This directory
1053 is searched before `data-dir'. It defaults to the same directory
1054 the actual main word list is in (which is not necessarily
1058 (dir) location for personal files
1061 (file) personal configuration file. This file overrides options
1062 found in the global `conf' file
1065 (file) use this keyboard layout for suggesting possible words.
1066 These spelling errors happen if a user accidently presses a key
1067 next to the intended correct key. The default is keyboard
1068 standard. If you are creating documents, you may want to set it
1069 according to your particular type of keyboard. If spellchecking
1070 documents created elsewhere, you might want to set this to the
1071 keyboard type for that locale. If you are not sure, just leave
1075 (dir) prefix directory
1078 (boolean) set the prefix based on executable location (only works
1079 on WIN32 and when compiled with `--enable-win32-relocatable')
1081 4.2.7 Aspell Utility Options
1082 ----------------------------
1085 (boolean) Create a backup file by appending `.bak' to the file
1086 name. This applies when the command is `check' and the backup
1087 file is only created if any spelling modifications take place.
1090 (boolean) Time load time and suggest time in `pipe' mode.
1093 (boolean) Use byte offsets instead of character offsets in `pipe'
1097 (boolean) Reverse the order of the suggestions list in `pipe' mode.
1100 (string) the keymapping to use. Either `aspell' for the default
1101 mapping or `ispell' to use the same mapping that the Ispell utility
1105 (boolean) make possible root/affix combinations not in the
1106 dictionary in `pipe' mode.
1109 (boolean) Suggest possible replacements in `pipe' mode. If false
1110 Aspell will simply report the misspelling and make no attempt at
1111 suggestions or possible corrections.
1114 File: aspell.info, Node: Dumping Configuration Values, Next: Notes on Various Options, Prev: The Options, Up: Customizing Aspell
1116 4.3 Dumping Configuration Values
1117 ================================
1119 To find out the current value of all the options use the command
1120 `aspell dump config'. This will dump the current Aspell configuration
1121 to standard output. The format of the contents dumped is such that it
1122 can be used as either the global or your personal configuration file.
1124 To find out the current value of a particular option use `aspell
1125 config OPTION'. This will print out the value of OPTION to `stdout'
1129 File: aspell.info, Node: Notes on Various Options, Prev: Dumping Configuration Values, Up: Customizing Aspell
1131 4.4 Notes on Various Options
1132 ============================
1136 * Notes on Various Filters and Filter Modes::
1137 * Notes on the Prefix Option::
1138 * Notes on Typo-Analysis::
1139 * Notes on the Different Suggestion Modes::
1142 File: aspell.info, Node: Notes on Various Filters and Filter Modes, Next: Notes on the Prefix Option, Up: Notes on Various Options
1144 4.4.1 Notes on Various Filters and Filter Modes
1145 -----------------------------------------------
1147 Aspell now has filter support. You can either select from individual
1148 filters or choose a filter mode. To select a filter mode use the
1149 `mode' option. You may choose from `none', `url', `email', `sgml',
1150 `ccpp', `tex' and any other available on your system. The default mode
1151 is `url'. Individual filters can be added with the option `add-filter'
1152 and removed with the `rem-filter' option. The currently available
1153 filters are `url', `email', `sgml' and `tex', `latex' (alias for
1154 `tex'), `nroff', `context', as well as a bunch of filters which
1155 translate the text from one format to another.
1157 To check which filters are available use `aspell dump filters'. To
1158 check which filter modes are available use `aspell dump modes'. The
1159 `aspell help' command will also list all available filter and filter
1162 4.4.1.1 None Filter Mode
1163 ........................
1165 The `none' mode is exactly what it says. It turns off all filters.
1170 The `url' filter/mode skips over URLs, host names, and email addresses.
1171 Because this filter is almost always useful and rarely does any harm
1172 it is enabled in all modes except `none'. To turn it off either select
1173 the `none' mode or use `rem-filter' option _after_ the desired mode is
1176 4.4.1.3 Email Filter
1177 ....................
1179 The `email' filter mode skips over quoted text. It currently does not
1180 support skipping over headers however a future version should. In the
1181 meantime I suggest you use Aspell with Newsbody which can be found at
1182 `http://home.worldonline.dk/~byrial/newsbody/'. The option
1183 `email-skip' controls the number of characters that can appear before
1184 the email quote character, the default is 10. The option
1185 `add|rem-email-quote' controls the characters that are considered quote
1186 characters, the defaults are `>' and `|'.
1191 The SGML filter allows you to spell check SGML, HTML, XHTML, and XML
1192 files. In most cases everything within a tag `<tag attrib=value
1193 attrib2="a whole sentence">' will be skipped by the spell checker. The
1194 SGML/HTML/XML that Aspell supports is a slight superset of most DTDs
1195 (Document Type Definitions) and can spell check the often non-conforming
1196 HTML found on the web.
1198 Two configuration options, `sgml-skip' and `sgml-check', allow you
1199 to control what is spell checked. The tag and attribute names specified
1200 are case insensitive.
1203 This is a list of tags whose contents will also be skipped by the
1204 spell checker. For example, if you wish to leave a misspelling in
1205 a document and not have them flagged as misspellings, you could
1206 surround them with a <nospellcheck> tag:
1207 <TD><FONT size=2><NOSPELLCHECK>leviosa</NOSPELLCHECK>
1208 is what Mr. Potter said</FONT></TD>
1209 And put that word in the skip config directive:
1210 add-sgml-skip nospellcheck
1213 This is a list of attributes whose values you do want spell
1214 checked. By default, 'alt' (<img> alternate text) is a member of
1215 the check list since it is text that is seen by a web page viewer.
1216 You may also want 'value' to be on the check list since that is
1217 the text put on buttons:
1218 add-sgml-check value
1219 In this case `<input type=button value="Donr">' will be flagged as
1222 This filter will also translate SGML characters of the form
1223 `&#num;'. Other SGML characters such as `&' will simply be skipped
1224 over so that the word `amp', for example, will not be spell checked.
1225 Eventually full support for properly translating SGML characters will
1231 The `html' filter is like the SGML Filter Mode but specialized for
1232 HTML. By default, 'script' and 'style' are members of the skip list in
1235 4.4.1.6 TeX/LaTeX Filter
1236 ........................
1238 The `tex' (all lowercase) filter mode skips over TeX commands and
1239 parameters and/or options to certain commands. It also skips over TeX
1240 comments by default. The option `[dont-]tex-check-comments' controls
1241 whether or not Aspell will skip over TeX comments. The option
1242 `add|rem-tex-command' controls which TeX commands should have certain
1243 parameters and/or options also skipped over. Commands that are not
1244 specified will have all their parameters and/or options checked. The
1245 format for each item is
1247 <command> <a list of p,P,o and Os>
1249 The first item is simply the command name. The second item controls
1250 which parameters to skip over. A 'p' skips over a parameter while a
1251 'P' doesn't. Similarly an 'o' will skip over an optional parameter
1252 while an 'O' doesn't. The first letter on the list will apply to the
1253 first parameter, the second letter will apply to the second parameter
1254 etc. If there are more parameters than letters Aspell will simply
1255 check them as normal. For example the option
1257 add-tex-command rule pp
1259 will skip over the first two parameters of the `rule' command while the
1262 add-tex-command foo Pop
1264 will _check_ the first parameter of the `foo' command, skip over the
1265 next optional parameter, if it is present, and will skip over the
1266 second parameter -- even if the optional parameter is not present --
1267 and will check any additional parameters.
1269 A `*' at the end of the command is simply ignored. For example the
1274 will ignore the first parameter in both `enlargethispage' and
1277 To remove a command simply use the `rem-tex-command' option. For
1282 will remove the command foo, if present, from the list of TeX commands.
1284 The TeX filter mode is also available via `latex' alias name.
1286 4.4.1.7 Texinfo Filter
1287 ......................
1289 The `texinfo' filter allows you to spell check Texinfo files. It will
1290 skip over any Texinfo commands and their parameters when appropriate.
1291 It will also skip over some Texinfo environments such as `example'.
1292 The list option `texinfo-ignore' controls which commands to ignore the
1293 parameters of and the list option `texinfo-ignore-env' controls which
1294 Texinfo environments to ignore.
1296 The Texinfo filter has special code to deal with the `@table' and
1297 related commands. It will apply the formatting command to each of the
1298 `@item' or `@itemx' commands just like Texinfo will. This means that
1299 if the formatting command is `@code' and and the `@code' command is a
1300 member of the `texinfo-ignore' option than the Texinfo filter will
1301 ignore the parameter of the `@item' command as if the parameter was also
1302 the parameter of the `@code' command.
1304 The Texinfo filter will also skip over the `\input texinfo' line.
1306 4.4.1.8 Nroff Filter
1307 ....................
1309 The `nroff' filter mode allows you to check the spelling of Nroff
1310 documents. The mode is enabled by giving `--add-filter=nroff' or `-n'
1311 command line option to `aspell'. It is also automatically enabled if
1312 the first three characters of the file being checked are `.\"' (a
1313 `nroff' comment marker) or the file name ends in a one of the following
1316 * single decimal digit from `0' to `9'
1322 This filter mode skips following `nroff' language elements:
1328 * Names of `nroff' registers (both traditional two-letter names and
1329 GNU nroff long names)
1331 * Arguments to the following requests: `ds', `de', `nr', `do', `so'.
1333 * Arguments to font switch (`\f') and size switch (`\s') escapes
1335 * Arguments to extended charset escape in both traditional (`\(')
1336 and extended (`\[comp1 comp2 ...]') form.
1338 4.4.1.9 Context Filter
1339 ......................
1341 The _context_ filter allows Aspell to distinguish between visible and
1342 invisible contexts. The visible ones will be spell checked and the
1343 invisible ones will be ignored. The contexts are distinguished by the
1344 fact that the visible/invisible ones are delimited by specific and
1345 unique delimiter characters or character sequences. Whether the
1346 delimited contexts should be visible or invisible only stated by the
1347 value of the `[dont-]context-visible-first' option and not by the
1350 The context delimiters are specified as pairs of delimiters via the
1351 `add|rem-context-delimiters' option. The delimiters enclosing a
1352 specific context are specified as a space separated pair. If more than
1353 one delimiter pair is specified by one call of
1354 `add|rem-context-delimiters' they have to be combined to a comma
1355 separated list. To indicate that a context is always closed by end of
1356 line use `\0' sequence as closing delimiter.
1358 4.4.1.10 Ccpp Filter Mode
1359 .........................
1361 The `ccpp' filter mode will limit spell checking to C/C++ comments and
1362 string literals. Any code in between will be left alone.
1365 File: aspell.info, Node: Notes on the Prefix Option, Next: Notes on Typo-Analysis, Prev: Notes on Various Filters and Filter Modes, Up: Notes on Various Options
1367 4.4.2 Notes on the Prefix Option
1368 --------------------------------
1370 The `prefix' option is there to allow Aspell to easily be relocated.
1371 Changing `prefix' will change all directory names relative to the new
1372 prefix that are not explicitly set. For example if `prefix' was
1373 `/usr/local/aspell' and `dict-dir' has a default value of
1374 `/usr/local/aspell/dict' than changing `prefix' to `/opt/aspell' will
1375 also change the default value of `dict-dir' to `/opt/aspell/dict'.
1376 Note that modifying `prefix' will only affect the default compiled in
1377 values of directories. If a directory option is explicitly given a
1378 value then changing the value of `prefix' has no effect on that
1382 File: aspell.info, Node: Notes on Typo-Analysis, Next: Notes on the Different Suggestion Modes, Prev: Notes on the Prefix Option, Up: Notes on Various Options
1384 4.4.3 Notes on Typo-Analysis and the Keyboard Definition File
1385 -------------------------------------------------------------
1387 Aspell .33 and better will, in general, give a higher priority to
1388 certain misspellings which are likely to be due to typos such as `teh'
1389 instead of `the' or `hapoy' instead of `happy'. However in order to do
1390 this well Aspell needs to know the layout of the keyboard via the
1391 keyboard definition file. The keyboard definition file simply
1392 identifies the keys on the keyboard and which of them are right next to
1393 each other. It has an extension of `.kbd' and all non-ASCII characters
1394 are expected to be in UTF-8.
1396 To identify a key use:
1399 where BASE is the base character that the key types, and OTHER are
1400 other keys that the key can produce. For example
1404 It generally is only necessary to list keys which type more than one
1405 distinct letter as Aspell can derive the rest from the language data
1406 file. For example, it is not necessary to include the previously
1409 To identify two keys as being right next to each other simply list
1410 the type keys right after each other. For example the line:
1414 will indicate that `a' and `s' are right next to each other. If `as'
1415 is listed as an entry it is not necessary to list `sa' as an entry as
1416 that will be done automatically. Also by "right next to each other" I
1417 mean two keys that are close enough together that it is easy to type
1418 one instead of the other. On most keyboards this means keys that are
1419 to the left or to the right of each other and _not_ keys that are below
1422 The default for this option is normally `standard'. However the
1423 default can be changed via the language data file. The normal default,
1424 `standard', should work well for most QWERTY like keyboard layouts. It
1425 may need minor adjusting for foreign keyboards. The `dvorak' option
1426 can be used for a Dvorak layout.
1429 File: aspell.info, Node: Notes on the Different Suggestion Modes, Prev: Notes on Typo-Analysis, Up: Notes on Various Options
1431 4.4.4 Notes on the Different Suggestion Modes
1432 ---------------------------------------------
1434 In order to understand what these suggestion modes do, a basic
1435 understanding of how Aspell works is required. For that, see *Note
1436 Aspell Suggestion Strategy::.
1438 The suggestion modes are as follows.
1441 This method will use the fastest method available to come up with
1442 decent suggestions. This currently means that it will look for
1443 soundslikes within one edit distance. This method will also use
1444 the replacement table if one is available. In this mode Aspell
1445 gets about 87% of the words from my small test kernel of
1446 misspelled words. (Go to `http://aspell.net/test' for more info
1447 on the test kernel as well as comparisons of this version of
1448 Aspell with previous versions and other spell checkers.)
1451 This method is currently identical to `ultra'.
1454 This mode will use what ever method is necessary to return good
1455 suggestions in most cases in a reasonable amount of time. This
1456 currently means it will looks for soundslikes within two edit
1457 distance apart. This mode gets 93% of the words.
1460 Like `normal' except that "reasonable amount of time" is not a
1461 consideration. In most cases it will return the same results as
1462 `normal'. The biggest difference is that it will try an ngram
1463 scan if the normal methods of finding a suggestion fail.
1466 This method is like `slow' but is tailored more for the bad
1467 speller, where as the other modes are tailored more to strike a
1468 good balance between typos and true misspellings. This mode never
1469 performs typo-analysis and returns a _huge_ number of words for
1470 the really bad spellers who can't seem to get the spelling
1471 anything close to what it should be. If the misspelled word looks
1472 anything like the correct spelling it is bound to be found
1473 _somewhere_ on the list of 100 or more suggestions. This mode
1474 gets 98% of the words.
1476 If jump tables were not used then the `normal' option is identical
1477 to `fast' and the `slow' option is identical to the `normal' if jump
1481 File: aspell.info, Node: Working With Dictionaries, Next: Writing programs to use Aspell, Prev: Customizing Aspell, Up: Top
1483 5 Working With Dictionaries
1484 ***************************
1488 * Using aspell-import::
1489 * How Aspell Selects an Appropriate Dictionary::
1490 * Listing Available Dictionaries::
1491 * Dumping the Contents of the Word List::
1492 * Creating an Individual Word List::
1493 * Working With Affix Info in Word Lists::
1494 * Format of the Personal and Replacement Dictionaries::
1495 * Using Multi Dictionaries::
1496 * Dictionary Naming::
1500 File: aspell.info, Node: Using aspell-import, Next: How Aspell Selects an Appropriate Dictionary, Up: Working With Dictionaries
1502 5.1 Using `aspell-import'
1503 =========================
1505 The `aspell-import' Perl script will look for old personal dictionaries
1506 and will import them into GNU Aspell. It will look for both Ispell and
1507 Aspell ones. To use it, just run it from the command prompt. If you
1508 get an error about `/usr/bin/perl' not being found, then instead try
1509 `perl BINDIR/aspell-import'. When running the script if you get a
1513 Error: No word lists can be found for the language "de".
1515 This means that you have not installed support for the given
1516 language, in this case `de' for German. To rectify the situation
1517 download and install a dictionary designed to work with GNU Aspell 0.50
1521 File: aspell.info, Node: How Aspell Selects an Appropriate Dictionary, Next: Listing Available Dictionaries, Prev: Using aspell-import, Up: Working With Dictionaries
1523 5.2 How Aspell Selects an Appropriate Dictionary
1524 ================================================
1526 If the `master' option is set in any fashion (via the command line, the
1527 `ASPELL_CONF' environment variable, or a configuration file) Aspell
1528 will look for a dictionary of that name. If one could not be found, it
1531 Otherwise it will use the value of the `lang' option to search for
1532 an appropriate dictionary. If more than one dictionary is found for
1533 the given language string then it will look for a dictionary with a
1534 matching variety if the `variety' option is set. If it is not set it
1535 will look for a dictionary without a variety. If after matching the
1536 `lang' and `variety' there is still more than one dictionary available
1537 it will find one with the size closest to the value of the `size'
1538 option. The default size is 60. If Aspell cannot find a dictionary
1539 based on the `lang' option then it will give up and complain.
1541 If the `lang' option is not explicitly set its value will be based
1542 on the `LC_MESSAGES' locale. This locale is generally taken from the
1543 `LC_MESSAGES' environment variable or the `LANG' environment variable
1544 if `LC_MESSAGES' is not set. However, if Aspell is being used as a
1545 library from within another program which already explicitly set the
1546 locale then it will use the locale of the library rather than the
1547 environment variables. If Aspell cannot determine the language from
1548 the `LC_MESSAGES' locale than it will default to `en_US'.
1550 The list option `dict-alias' can be used to influence which
1551 dictionary is selected by creating an alias from one dictionary name to
1552 another. This option is most useful when there is more than one
1553 dictionary for a given language. For example `add-dict-alias en_US
1554 en_US-w_accents' will cause Aspell to choose the accented version of
1555 the American English dictionary instead of the non-accented version.
1556 To add an alias use:
1558 add-dict-alias NAME VAL
1561 File: aspell.info, Node: Listing Available Dictionaries, Next: Dumping the Contents of the Word List, Prev: How Aspell Selects an Appropriate Dictionary, Up: Working With Dictionaries
1563 5.3 Listing Available Dictionaries
1564 ==================================
1566 For a list of available dictionaries use the command `aspell dump
1567 dicts'. This will form a list of dictionaries that Aspell will search
1568 when a dictionary is not specifically given.
1571 File: aspell.info, Node: Dumping the Contents of the Word List, Next: Creating an Individual Word List, Prev: Listing Available Dictionaries, Up: Working With Dictionaries
1573 5.4 Dumping the Contents of the Word List
1574 =========================================
1576 The dump command in `aspell' will simply dump the contents of a word
1577 list to `stdout' in a format that can be read back in with `aspell
1580 If no word list is specified the command will act on the default one.
1581 For example the command
1583 aspell dump personal
1585 will simply dump the contents of the current personal word list to
1589 File: aspell.info, Node: Creating an Individual Word List, Next: Working With Affix Info in Word Lists, Prev: Dumping the Contents of the Word List, Up: Working With Dictionaries
1591 5.5 Creating an Individual Word List
1592 ====================================
1594 To create an individual main word list from a list of words use the
1597 aspell --lang=LANG create master ./BASE < WORDLIST
1599 where BASE is the name of the word list and WORDLIST is the list of
1600 words separated by white space. The name of the word list will
1601 automatically be converted to all lowercase. The `./' is important
1602 because without it Aspell will create the word list in the normal word
1603 list directory. If you are trying to create a word list in a language
1604 other than English check the Aspell `data-dir' (usually
1605 `/usr/share/aspell', use `aspell dump config' to find out what it is on
1606 your system) to see if a language data file exists for your language.
1607 If not you will need to create one. For more information on using
1608 Aspell with other languages *Note Adding Support For Other Languages::.
1610 This will create the file `BASE' in the current directory. To use
1611 the new word list copy the file to the normal word list directory (use
1612 `aspell config' to find out what it is) and use the option
1615 During the creating of the dictionary you may get a number of
1616 warnings or errors about invalid words or affixes. By default Aspell
1617 will skip any invalid words and remove invalid affixes. If you rather
1618 that Aspell simply accepts all words given then the option
1619 `--dont-validate-words' can be specified. To avoid checking if affixes
1620 are valid use the option `--dont-validate-affixes'. However, rather
1621 than disable checking, it is preferable to clean the input word list.
1622 This can be done by using the command
1624 aspell --local-data-dir=./ --lang=LANG clean < WORDLIST > RESULT
1626 which will clean the word list and output the results to RESULT. By
1627 default it will remove invalid characters from the beginning and end of
1628 a word before resorting to skipping the word. If you rather it just
1629 skip the words than add the keyword strict:
1631 aspell --local-data-dir=./ --lang=LANG clean strict < WORDLIST > RESULT
1633 The option `--clean-words' can be be added when creating a
1634 dictionary if you want Aspell to remove invalid characters from the
1635 beginning and end of a word like the "clean" command does. In addition
1636 the options `--dont-skip-invalid-words' and `--dont-clean-affixes' can
1637 be specified to turn the warnings into errors.
1639 The compiled dictionary file are endian order dependent. When a
1640 dictionary is loaded the endian order is checked. Please do not
1641 distribute the compiled dictionaries unless you are only distributing
1642 them for a particular platform such as you would a binary.
1644 Aspell is now also able to use special `multi' dictionaries. for
1645 more information *Note How Aspell Selects an Appropriate Dictionary::.
1647 A personal and replacement word list can be created in a similar
1650 5.5.1 Format of the Replacement Word List
1651 -----------------------------------------
1653 The replacement word list has each replacement pair on its own line in
1654 the following format
1656 misspelled_word correction
1659 File: aspell.info, Node: Working With Affix Info in Word Lists, Next: Format of the Personal and Replacement Dictionaries, Prev: Creating an Individual Word List, Up: Working With Dictionaries
1661 5.6 Working With Affix Info in Word Lists
1662 =========================================
1664 5.6.1 The Munch Command
1665 -----------------------
1667 The `munch' command takes a list of words from standard input and
1668 outputs a list of possible root words and affixes. The root may,
1669 however, be invalid as it does not check them against the existing
1670 dictionary. For example the command:
1671 echo brother | aspell -l en munch
1673 brother broth/R brothe/R
1675 5.6.2 The Expand Command
1676 ------------------------
1678 The `expand' command is the reverse of `munch', it expands affix flags
1679 to produce a list of words. For example:
1680 echo both/R | aspell -l en expand
1684 The formal usage is:
1685 aspell expand [LEVEL] [LIMIT]
1686 Where LEVEL is the expansion level. Valid values are between 1 and
1687 3. Level 1 is the default if not otherwise specified. Level 2 causes
1688 the original root/affix to be included, for example:
1690 Level 3 causes multiple lines to be printed, one for each generated
1691 word, with the original root/affix combination followed by the word it
1695 Levels larger than 3 may also be supported, but should not be used as
1696 they may eventually be removed.
1698 If a LIMIT parameter is given then only expansions which affect the
1699 first LIMIT letters will be expanded. If a base word is not completely
1700 expanded for a given affix flag that flag will be left on the word.
1701 Note that prefixes are always expanded.
1703 5.6.3 The Munch-list Command
1704 ----------------------------
1706 The `munch-list' command will reduce the size of word list via affix
1707 compression. It will reduce a list of words to a minimal (or close to
1708 it) set of roots and affixes that will match the same list of words.
1709 The list of words is read from standard input and the result, the
1710 "munched" list, is written to standard out. It's usage is:
1712 aspell munch-list [keep] [single|multi] [simple] < INFILE > OUTFILE
1714 where `simple', `single', `multi', and `keep' are literal values.
1716 The default algorithm used should give near optimum results. In some
1717 cases the set of words returned is, provably, the minimum number
1718 possible. In the typical case the number of words returned is within
1719 1% of the optimum number.
1721 By default Aspell will remove redundant affix flags. The `keep'
1722 flag will avoid removing them, which can be useful if you want to
1723 include all possible expansions for each base word.
1725 When cross products are involved it may be beneficial to list a base
1726 word more than once. Unfortunately, the current version of Aspell can
1727 not correctly handle multiple base words in a dictionary. Therefore,
1728 the current default behavior is to only include the one with the most
1729 expansions. All of them can be included via the `multi' flag. Once
1730 Aspell is able to handle multiple base words the default will be to
1731 include them all. The `single' flag can be used to only include one of
1734 The `simple' flag will select an alternate faster algorithm. This
1735 algorithm is very similar to the `munch' command distributed with
1736 MySpell (the Open Office spell checker), however, it doesn't give
1737 nearly as good results. It does okay for the English word list but not
1738 for some other languages such as German; the normal algorithm reduced a
1739 list of 312,002 German words to 79,420 base words while the simple
1740 algorithm only reduced it to 115,927 words. This algorithm may
1741 disappear in a future version of Aspell.
1744 File: aspell.info, Node: Format of the Personal and Replacement Dictionaries, Next: Using Multi Dictionaries, Prev: Working With Affix Info in Word Lists, Up: Working With Dictionaries
1746 5.7 Format of the Personal and Replacement Dictionaries
1747 =======================================================
1749 5.7.1 Format of the Personal Dictionary
1750 ---------------------------------------
1752 The personal dictionary generally has a filename of the form:
1756 And the file itself contains two parts. The first part is a header
1759 personal_ws-1.1 LANG NUM [ENCODING]
1761 where NUM is the number of words in the list. This number is only used
1762 as a hint, and thus does not have to be accurate. When creating a new
1763 dictionary it is perfectly acceptable for NUM to be 0. The ENCODING is
1764 optional and specifies the encoding of the word list. If it is left
1765 out the encoding is expected to be in the default encoding for the
1766 language as specified by the `data-encoding' option. *Note
1769 The second part of simply a word list with one word per line.
1771 5.7.2 Format of the Personal Replacement Dictionary
1772 ---------------------------------------------------
1774 The personal replacement dictionary generally has a filename of the
1779 And the file itself contains two parts. The first part is a header
1782 personal_repl-1.1 LANG NUM [ENCODING]
1784 where NUM is currently unused and thus always 0. As with the personal
1785 dictionary the ENCODING is optional.
1787 The second part simply a list of replacements with one replacement
1788 per line with each replacement pair has the following format:
1790 MISSPELLED_WORD CORRECTION
1793 File: aspell.info, Node: Using Multi Dictionaries, Next: Dictionary Naming, Prev: Format of the Personal and Replacement Dictionaries, Up: Working With Dictionaries
1795 5.8 Using Multi Dictionaries
1796 ============================
1798 As with previous versions of Aspell you can specify the main dictionary
1799 to use via the `-d' or `--master' option. However as of Aspell .32 you
1802 1. Specify more than word list to use with the `extra-dicts' option.
1804 2. Specify special _multi_ dictionaries.
1806 The `extra-dicts' is a list option. To add a dictionary use
1807 `add-extra-dicts' or to remove a dictionary from the list use
1810 A _multi_ dictionary is a special file which is basically a list of
1811 dictionary files to use. A _multi_ dictionary must end in `.multi' and
1812 has roughly the same format as a configuration file with the only
1813 accepted key being `add'.
1816 File: aspell.info, Node: Dictionary Naming, Next: AWLI files, Prev: Using Multi Dictionaries, Up: Working With Dictionaries
1818 5.9 Dictionary Naming
1819 =====================
1821 In order for Aspell to be able to correctly recognize a dictionary
1822 based on the setting of the `LANG' environment variable the
1823 dictionaries need to be located somewhere Aspell can find them and they
1824 need to be _multi_ dictionaries. Where Aspell looks for dictionaries
1825 depends on the value of the `dict-dir' and `word-list-path' option.
1827 `dict-dir' is generally `PREFIX/lib/aspell', and `word-list-path' is
1830 Each dictionary that you expect Aspell to be able to find needs to
1831 have a name in the following format:
1833 LANGUAGE[_REGION][-VARIETY][-SIZE].multi
1835 where LANGUAGE is the two letter language code, REGION is the two
1836 letter region code, VARIETY is any extra information to distinguish the
1837 word list from other ones with the same language and spelling.
1838 Multiple varieties can be used by separating them with a '-'. Finally,
1839 SIZE is the size of the dictionary. If no size is specified then the
1840 default size of 60 will be assumed.
1852 File: aspell.info, Node: AWLI files, Prev: Dictionary Naming, Up: Working With Dictionaries
1857 In order for Aspell to find dictionaries that are located in odd places
1858 or not named according to *Note Dictionary Naming::, an AWLI file needs
1859 to be created for the dictionary and located in some place where Aspell
1862 Each AWLI file has a name in the following format:
1864 LANGUAGE[REGION][-VARIETY][-SIZE]-MODULE.awli
1866 where the names have the same meaning as in *Note Dictionary Naming::,
1867 and MODULE is the speller module to use, which should be set to DEFAULT
1868 for now since there is only one speller module.
1870 Each `awli' file for an Aspell word list should then contain exactly
1871 one line which contains the full path of the main word list.
1874 File: aspell.info, Node: Writing programs to use Aspell, Next: Adding Support For Other Languages, Prev: Working With Dictionaries, Up: Top
1876 6 Writing programs to use Aspell
1877 ********************************
1879 There are two main ways to use Aspell from within your application.
1880 Through the external C API or through a pipe. The internal Aspell API
1881 can be used directly but that is not recommended as the actual Aspell
1882 API is constantly changing.
1886 * Through the C API::
1888 * Notes on Storing Replacement Pairs::
1891 File: aspell.info, Node: Through the C API, Next: Through A Pipe, Up: Writing programs to use Aspell
1893 6.1 Through the C API
1894 =====================
1896 The Aspell library contains two main classes and several helper
1897 classes. The two main classes are `AspellConfig' and `AspellSpeller'.
1898 The `AspellConfig' class is used to set initial defaults and to change
1899 spell checker specific options. The `AspellSpeller' class does most of
1900 the real work. The `C API' is responsible for managing the
1901 dictionaries, checking if a word is in the dictionary, and coming up
1902 with suggestions among other things. There are many helper classes the
1903 important ones are `AspellWordList', `AspellMutableWordList',
1904 `Aspell*Enumeration'. The `AspellWordList' classes is used for
1905 accessing the suggestion list, as well as the personal and suggestion
1906 word list currently in use. The `AspellMutableWordList' is used to
1907 manage the personal, and perhaps other, word lists. The
1908 `Aspell*Enumeration' classes are used for iterating through a list.
1913 To use Aspell your application should include `aspell.h'. In order to
1914 ensure that all the necessary libraries are linked in libtool should be
1915 used to perform the linking. When using libtool simply linking with
1916 `-laspell' should be all that is necessary. When using shared
1917 libraries you might be able to simply link `-laspell', but this is not
1918 recommended. This version of Aspell uses the CVS version of libtool
1919 however released versions of libtool should also work.
1921 When your application first starts you should get a new configuration
1922 class with the command:
1924 AspellConfig * spell_config = new_aspell_config();
1926 which will create a new `AspellConfig' class. It is allocated with
1927 `new' and it is your responsibility to delete it with
1928 `delete_aspell_config'. Once you have the config class you should set
1929 some variables. The most important one is the language variable. To
1930 do so use the command:
1932 aspell_config_replace(spell_config, "lang", "en_US");
1934 which will set the default language to use to American English. The
1935 language is expected to be the standard two letter ISO 639 language
1936 code, with an optional two letter ISO 3166 country code after an
1937 underscore. You can set the preferred size via the `size' option, any
1938 extra info via the `variety' option, and the encoding via the
1939 `encoding' option. Other things you might want to set is the preferred
1940 spell checker to use, the search path for dictionaries, and the like --
1941 see *Note The Options::, for a list of all available options.
1943 Whenever a new document is created a new `AspellSpeller' class
1944 should also be created. There should be one speller class per
1945 document. To create a new speller class use the `new_aspell_speller'
1946 and then cast it up using `to_aspell_speller' like so:
1948 AspellCanHaveError * possible_err = new_aspell_speller(spell_config);
1949 AspellSpeller * spell_checker = 0;
1950 if (aspell_error_number(possible_err) != 0)
1951 puts(aspell_error_message(possible_err));
1953 spell_checker = to_aspell_speller(possible_err);
1955 which will create a new `AspellSpeller' class using the defaults found
1956 in `spell_config'. To find out which dictionary is selected the
1957 `lang', `size', and `variety' options may be examined. To find out the
1958 exact name of the dictionary the `master' option may be examined as
1959 well as the `master-flags' options to see if there were any special
1960 flags that were passed on to the module. The `module' option way also
1961 be examined to figure out which speller module was selected, but since
1962 there is only one this option will always be the same.
1964 If for some reason you want to use different defaults simply clone
1965 `spell_config' and change the setting like so:
1967 AspellConfig * spell_config2 = aspell_config_clone(spell_config);
1968 aspell_config_replace(spell_config2, "lang","nl");
1969 possible_err = new_aspell_speller(spell_config2);
1970 delete_aspell_config(spell_config2);
1972 Once the speller class is created you can use the `check' method to
1973 see if a word in the document is correct like so:
1975 int correct = aspell_speller_check(spell_checker, WORD, SIZE);
1977 WORD is expected to be a `const char *' character string. If the
1978 encoding is set to be `ucs-2' or `ucs-4' WORD is expected to be a cast
1979 from either `const u16int *' or `const u32int *' respectively.
1980 `u16int' and `u32int' are generally `unsigned short' and `unsigned int'
1981 respectively. SIZE is the length of the string or `-1' if the string
1982 is null terminated. If the string is a cast from `const u16int *' or
1983 `const u32int *' then `size' is the amount of space in bytes the string
1984 takes up after being cast to `const char *' and not the true size of
1985 the string. `sspell_speller_check' will return `0' if it is not found
1986 and non-zero otherwise.
1988 If the word is not correct, then the `suggest' method can be used to
1989 come up with likely replacements.
1991 AspellWordList * suggestions = aspell_speller_suggest(spell_checker,
1993 AspellStringEnumeration * elements = aspell_word_list_elements(suggestions);
1995 while ( (word = aspell_string_enumeration_next(aspell_elements)) != NULL )
1997 // add to suggestion list
1999 delete_aspell_string_enumeration(elements);
2001 Notice how `elements' is deleted but `suggestions' is not. The
2002 value returned by `suggestions' is only valid to the next call to
2003 `suggest'. Once a replacement is made the `store_repl' method should
2004 be used to communicate the replacement pair back to the spell checker
2005 (for the reason, *note Notes on Storing Replacement Pairs::). Its
2006 usage is as follows:
2008 aspell_speller_store_repl(spell_checker, MISSPELLED_WORD, SIZE,
2009 CORRECTLY_SPELLED_WORD, SIZE);
2011 If the user decided to add the word to the session or personal
2012 dictionary the the word can be be added using the `add_to_session' or
2013 `add_to_personal' methods respectively like so:
2015 aspell_speller_add_to_session|personal(spell_checker, word, size);
2017 It is better to let the spell checker manage these words rather than
2018 doing it yourself so that the words have a chance of appearing in the
2021 Finally, when the document is closed the `AspellSpeller' class
2022 should be deleted like so:
2024 delete_aspell_speller(spell_checker);
2029 Methods that return a boolean result generally return `false' on error
2030 and `true' otherwise. To find out what went wrong use the
2031 `error_number' and `error_message' methods. Unless otherwise stated
2032 methods that return a `const char *' will return `NULL' on error. In
2033 general, the character string returned is only valid until the next
2034 method which returns a `const char *' is called.
2036 For the details of the various classes please see the header files.
2037 In the future I will generate class references using some automated
2043 Two simple examples are included in the examples directory. The
2044 `example-c' program demonstrates most of the Aspell library
2045 functionality and the `list-dicts' lists the available dictionaries.
2047 6.1.4 Notes About Thread Safety
2048 -------------------------------
2050 Aspell should be thread safe, when used properly, as long as the
2051 underlying compiler, C and C++ library is thread safe. Aspell objects,
2052 including the AspellSpeller class, should not be used by multiple
2053 threads unless they are protected by locks or it is only accessed by
2054 read-only methods. A method is read-only only if a `const' object is
2055 passed in. Many methods that seam to be read-only are not because they
2056 may store state information in the object.
2059 File: aspell.info, Node: Through A Pipe, Next: Notes on Storing Replacement Pairs, Prev: Through the C API, Up: Writing programs to use Aspell
2064 When given the `pipe' or `-a' command, Aspell goes into a pipe mode
2065 that is compatible with `ispell -a'. Aspell also defines its own set
2066 of extensions to Ispell pipe mode.
2068 6.2.1 Format of the Data Stream
2069 -------------------------------
2071 In this mode, Aspell prints a one-line version identification message,
2072 and then begins reading lines of input. For each input line, a single
2073 line is written to the standard output for each word checked for
2074 spelling on the line. If the word was found in the main dictionary, or
2075 your personal dictionary, then the line contains only a `*'.
2077 If the word is not in the dictionary, but there are suggestions, then
2078 the line contains an `&', a space, the misspelled word, a space, the
2079 number of near misses, the number of characters between the beginning
2080 of the line and the beginning of the misspelled word, a colon, another
2081 space, and a list of the suggestions separated by commas and spaces.
2083 If you set the option `run-together' and Aspell thinks this word is
2084 a combination of two words in the dictionary, then it prints a single
2087 Finally, if the word does not appear in the dictionary, and there are
2088 no suggestions, then the line contains a `#', a space, the misspelled
2089 word, a space, and the character offset from the beginning of the line.
2090 Each sentence of text input is terminated with an additional blank
2091 line, indicating that Aspell has completed processing the input line.
2093 These output lines can be summarized as follows:
2096 *Suggestions*: & original count offset: miss, miss, ...
2097 *None*: # original offset
2099 When in the `-a' mode, Aspell will also accept lines of single words
2100 prefixed with any of `*', `&', `@', `+', `-', `~', `#', `!', `%', or
2101 `^'. A line starting with `*' tells Aspell to insert the word into the
2102 user's dictionary. A line starting with `&' tells Aspell to insert an
2103 all-lowercase version of the word into the user's dictionary. A line
2104 starting with `@' causes Aspell to accept this word in the future. A
2105 line starting with `+', followed immediately by a valid mode will cause
2106 Aspell to parse future input according the syntax of that formatter. A
2107 line consisting solely of a `+' will place Aspell in TeX/LaTeX mode
2108 (similar to the `-t' option) and `-' returns Aspell to its default mode
2109 (which is Nroff unless otherwise specified). (but these commands are
2110 obsolete). A line `~', is ignored for Ispell compatibility. A line
2111 prefixed with `#' will cause the personal dictionaries to be saved. A
2112 line prefixed with `!' will turn on terse mode (see below), and a line
2113 prefixed with `%' will return Aspell to normal (non-terse) mode. Any
2114 input following the prefix characters `+', `-', `#', `!', `~', or `%'
2115 is ignored, as is any input following. To allow spell-checking of
2116 lines beginning with these characters, a line starting with `^' has
2117 that character removed before it is passed to the spell-checking code.
2118 It is recommended that programmatic interfaces prefix every data line
2119 with an uparrow to protect themselves against future changes in Aspell.
2123 `*WORD' Add a word to the personal dictionary
2124 `&WORD' Insert the all-lowercase version of the word in the personal
2126 `@WORD' Accept the word, but leave it out of the dictionary
2127 `#' Save the current personal dictionary
2128 `~' Ignored for Ispell compatibility.
2130 `+MODE' Enter the mode specified by MODE.
2131 `-' Enter the default mode.
2132 `!' Enter terse mode
2134 `^' Spell-check the rest of the line
2136 In terse mode, Aspell will not print lines beginning with `*', which
2137 indicate correct words. This significantly improves running speed when
2138 the driving program is going to ignore correct words anyway.
2140 In addition to the above commands which are designed for Ispell
2141 compatibility Aspell also supports its own extensions. All Aspell
2142 extensions follow the following format.
2146 Where DATA may or may not be required depending on the particular
2147 command. Aspell currently supports the following commands:
2149 `cs OPTION,VALUE' Change a configuration option.
2150 `cr OPTION' Prints the value of a configuration option.
2151 `pp' Returns a list of all words in the current
2153 `ps' Returns a list of all words in the current
2155 `l' Returns the current language name.
2156 `ra MIS,COR' Add the word pair to the replacement
2157 dictionary for later use. Returns nothing.
2159 Anything returned is returned on its own line. All lists returned
2160 have the following format
2162 num of items: item1, item2, etc
2164 _(Part of the preceding section was directly copied out of the
2168 File: aspell.info, Node: Notes on Storing Replacement Pairs, Prev: Through A Pipe, Up: Writing programs to use Aspell
2170 6.3 Notes on Storing Replacement Pairs
2171 ======================================
2173 The `store_repl' method and the `$$ra' should be used because Aspell is
2174 able to learn from users misspellings. For example on the first pass a
2175 user misspells _beginning_ as _beging_ so Aspell suggests:
2177 begging, begin, being, Beijing, bagging, ....
2179 However the user then tries _begning_ and Aspell suggests
2181 beginning, beaning, begging, ...
2183 so the user selects _beginning_. However, later on in the document the
2184 user misspells it as _begng_ (*not* _beging_). Normally Aspell will
2187 began, begging, begin, begun, ...
2189 However because it knows the user misspelled _beginning_ as _beging_ it
2190 will instead suggest:
2192 beginning, began, begging, begin, begun ...
2194 I myself often misspelled beginning (and still do) as something close
2195 to begging and too many times wind up writing sentences such as
2198 Please also note that replacements commands have a memory. Which
2199 means if you first store the replacement pair:
2201 sicolagest -> psycolagest
2203 then store the replacement pair
2205 psycolagest -> psychologist
2207 The replacement pair
2209 sicolagest -> psychologist
2211 will also get stored so that you don't have to worry about it.
2214 File: aspell.info, Node: Adding Support For Other Languages, Next: Implementation Notes, Prev: Writing programs to use Aspell, Up: Top
2216 7 Adding Support For Other Languages
2217 ************************************
2219 Before you consider adding support for Aspell, first make sure that
2220 someone else has not already done it. A good number of dictionaries
2221 off the Aspell home page at `http://aspell.net'. If your language is
2222 not listed above feel free to send mail to aspell-dict at gnu org for
2223 help in getting started.
2225 Adding a language to Aspell is fairly straightforward. You basically
2226 need to create the language data file, and compile a new word list.
2230 * The Language Data File::
2231 * Compiling the Word List::
2233 * The Simple Soundslike::
2234 * Replacement Tables::
2235 * Affix Compression::
2236 * Controlling the Behavior of Run-together Words::
2237 * Creating A New Character Set::
2238 * Creating An Official Dictionary Package::
2241 File: aspell.info, Node: The Language Data File, Next: Compiling the Word List, Up: Adding Support For Other Languages
2243 7.1 The Language Data File
2244 ==========================
2246 The basic format of the language data file is the same as it is for the
2247 Aspell configuration file. It is named `LANG.dat' and is located in
2248 the architecture independent data dir for Aspell (option `data-dir')
2249 which is usually `PREFIX/share/aspell'. Use `aspell config' to find
2250 out where it is in your installation. By convention the language name
2251 should be the two letter ISO 639 language code if it exists, if not use
2252 the three letter code.
2254 The language data file has several mandatory fields, and several
2255 optional ones. All fields are case sensitive and should be in all
2258 The two mandatory fields are `name' and `charset'.
2260 `name' is the name of the language and should be the same as the
2261 file name (without the `.dat').
2263 `charset' is the 8-bit character set Aspell will expect the word
2264 lists to be formatted in. If possible choose from one of the standard
2265 ones provided with Aspell. These are `iso-8859-*', `koi8-*', or
2266 `viscii'. If your language does not require any non-ascii characters
2267 choose `iso-8859-1'. If one of these standard character sets is not
2268 suitable for your language then you can create a new one. *Note
2269 Creating A New Character Set::.
2271 The optional fields are as follows:
2274 The encoding the language data files are expected to be in as well
2275 as the default encoding to use when saving the personal
2276 dictionaries. It can be either `utf-8' or any of the 8-bit
2277 encoding that Aspell supports. If not set, then it defaults to
2281 Non-letter characters that can appear in your language such as the
2282 `'' and `-'. The format for the value is a list separated by
2283 spaces. Each item of the list has the following format.
2285 <char> <begin><middle><end>
2287 CHAR is the non-letter character in question. BEGIN, MIDDLE, END
2288 are either a `-' or a `*'. A star for BEGIN means that the
2289 character can begin a word, a `-' means it can't. The same is
2290 true for MIDDLE and END. For example, the entry for the `'' in
2295 To include more than one middle character just list them one after
2296 another on the same line. For example, to make both the `'' and
2297 the `-' a middle character, use the following line in the language
2302 However, please be aware that adding special characters can have
2303 unintended consequences due to limitations of Aspell. For example
2304 if the `-' was accepted as a middle character, then _every_ word
2305 with a `-' in it would be flagged as a spelling error unless that
2306 exact word is in the dictionary, even if both parts are in the
2307 dictionary. Also, having a `.' as an end character will cause the
2308 `.' to be part of any misspelled words. Which can get very
2309 annoying if you misspell a word at the end of a sentence.
2312 The name of the soundslike data for the language. The data is
2313 expected to be in the file `NAME_phonet.dat'.
2315 If NAME is `simpile' then a very simple soundslike is used. This
2316 is not as powerful as full phonetic soundslike but it can be
2317 computed a lot faster. (*note The Simple Soundslike::)
2319 If the soundslike name is `none', or this option is not specified,
2320 then no soundslike will be used. The effective soundslike is the
2321 word converted to all lowercase and possibly with accents stripped
2322 depending on the `store-as' option. For languages with phonetic
2323 spelling the difference will not be very noticeable. However, for
2324 languages with non-phonetic spelling there will be a noticeable
2325 difference. The difference you notice will depend on the quality
2326 of the soundslike data file. If you do not notice much of a
2327 difference for a language with non-phonetic spelling that is a good
2328 indication that the soundslike data is not rough enough--or the
2329 words you are trying are not that badly misspelled.
2331 `invisible-soundslike'
2332 Avoid storing the soundslike information with the word. Instead
2333 it is computed as needed. This option defaults to true if the
2334 soundslike is `none' or `simpile', and false when a phonetic
2338 *Note Replacement Tables::.
2341 The base name of the keyboard definition file to use. For more
2342 information see *Note Notes on Typo-Analysis::.
2345 A list of characters which specifies which characters to insert
2346 between two words when a word is split. This is a list option.
2351 *Note Affix Compression::.
2354 How the words are indexed in the dictionary. If "stripped" then
2355 the word is indexed in a lower case and de-accented form. If
2356 "lower", then the word is indexed in a lower case form but with
2357 accent info still intact. This just controls how the word is
2358 indexed, not how it is stored. The default is "stripped" unless
2359 affix compression is used.
2362 Should be set to true if your language makes use of private use
2363 characters or when Normalization Form C is not the same as full
2370 Additional options includes options to control how run-together words
2371 are handled the same way as they are in the normal configuration files.
2372 for more information, please *Note Controlling the Behavior of
2373 Run-together Words::.
2376 File: aspell.info, Node: Compiling the Word List, Next: Phonetic Code, Prev: The Language Data File, Up: Adding Support For Other Languages
2378 7.2 Compiling the Word List
2379 ===========================
2381 Once you have a working language data file installed in the right place
2382 you are ready to compile the main word list. To find out what to do,
2383 see *Note Working With Dictionaries::. This section also includes
2384 instructions for creating the AWLI file.
2387 File: aspell.info, Node: Phonetic Code, Next: The Simple Soundslike, Prev: Compiling the Word List, Up: Adding Support For Other Languages
2392 Aspell is in fact the spell checker that comes up with the best
2393 suggestions if it finds an unknown word. One reason is that it does
2394 not just compare the word with other words in the dictionary (like
2395 Ispell does) but also uses phonetic comparisons with other words.
2397 The new table driven phonetic code is very flexible and setting up
2398 phonetic transformation rules for other languages is not difficult but
2399 there can be a number of stumbling blocks -- that's why I wrote this
2402 The main phonetic code is free of any language specific code and
2403 should be powerful enough to allow setting up rules for any language.
2404 Anything which is language specific is kept in a plain text file and
2405 can easily be edited. So it's even possible to write phonetic
2406 transformation rules if you don't have any programming skills. All you
2407 need to know is how words of the language are written and how they are
2410 7.3.1 Syntax of the transformation array
2411 ----------------------------------------
2413 In the translation array there are two strings on each line; the first
2414 one is the search string (or switch name) and the second one is the
2415 replacement string (or switch parameter). The line
2419 is also required to appear somewhere in the translation array. The
2420 version string can be anything but it should be changed whenever a new
2421 version of the translation array is released. This is important
2422 because it will keep Aspell from using a compiled dictionary with the
2423 wrong set of rules. For example, if when coming up with suggestion for
2424 `hallo', Aspell will use the new rules to come up with the soundslike
2425 say `H*L*', but if `hello' is stored in the dictionary using the old
2426 rules as `HL' instead of `H*L*' Aspell will never be able to come up
2427 with `hello'. So to solve this problem Aspell checks if the version
2428 strings match and aborts with an error if they don't. Thus it is
2429 important to update it whenever a new version of the translation array
2430 is released. This is only a problem with the main word list as the
2431 personal word lists are now stored as simple word lists with a single
2432 header line (i.e. no soundslike data).
2434 Each non switch line represents one replacement (transformation)
2435 rule. Words beginning with the same letter must be grouped together;
2436 the order inside this group does not depend on alphabetical issues but
2437 it gives priorities; the higher the rule the higher the priority.
2438 That's why the first rule that matches is applied. In the following
2444 `GH -> _' has higher priority than `G -> K'
2446 `_' represents the empty string "". If `GH -> _' came after `G ->
2447 K', the second rule would never match because the algorithm would stop
2448 searching for more rules after the first match. The above rules
2449 transform any `GH' to an empty string (delete them) and transforms any
2452 At the end of the first string of a line (the search string) there
2453 may optionally stand a number of characters in brackets. One (only
2454 one!) of these characters must fit. It's comparable with the `[ ]'
2455 brackets in regular expressions. The rule `DG(EIY) -> J' for example
2456 would match any `DGE', `DGI' and `DGY' and replace them with `J'. This
2457 way you can reduce several rules to one.
2459 Before the search string, one or more dashes `-' may be placed.
2460 Those search strings will be matched totally but only the beginning of
2461 the string will be replaced. Furthermore, for these rules no follow-up
2462 rule will be searched (what this is will be explained later). The rule
2463 `TCH-- '-> _ will match any word containing `TCH' (like `match') but
2464 will only replace the first character `T' with an empty string. The
2465 number of dashes determines how many characters from the end will not
2466 be replaced. After the replacement, the search for transformation
2467 rules continues with the not replaced `CH'!
2469 If a `<' is appended to the search string, the search for
2470 replacement rules will continue with the replacement string and not with
2471 the next character of the word. The rule `PH< -> F' for example would
2472 replace `PH' with `F' and then again start to search for a replacement
2473 rule for `F...'. If there would also be rules like `FO '-> `O' and `F
2474 -> _' then words like `PHOXYZ' would be transformed to `OXYZ' and any
2475 occurrences of `PH' that are not followed by an `O' will be deleted like
2476 `PHIXYZ -> IXYZ'. The second replacement however is not applied if the
2477 priority of this rule is lower than the priority of the first rule.
2479 Priorities are added to a rule by putting a number between 0 and 9 at
2480 the end of the search string, for example `ING6 -> N'. The higher the
2481 number the higher is the priority.
2483 Priorities are especially important for the previously mentioned
2484 follow-up rules. Follow-up rules are searched beginning from the last
2485 string of the first search string. This is a bit complicated but I
2486 hope this example will make it clearer:
2495 In this example `CHS' in the word `FUCHS' would be transformed to
2496 `X'. If we take the word `DURCHSCHNITT' then things look a bit
2497 different. Here `CH' belongs together and `SCH' belongs together and
2498 both are spoken separately. The algorithm however first finds the
2499 string `CHS' which may not be transformed like in the previous word
2500 `FUCHS'. At this point the algorithm can find a follow-up rule. It
2501 takes the last character of the first matching rule (`CHS') which is
2502 `S' and looks for the next match, beginning from this character. What
2503 it finds is clear: It finds `SCH -> SH', which has the same priority
2504 (no priority means standard priority, which is 5). If the priority is
2505 the same or higher the follow-up rule will be applied. Let's take a
2506 look at the word `SCHAUKEL'. In this word `SCH' belongs together and
2507 may not be taken apart. After the algorithm has found `SCH '-> `SH' it
2508 searches for a follow-up rule for `H+'`AUKEL'. It finds `HAU--1 -> H',
2509 but does not apply it because its priority is lower than the one of the
2510 first rule. You see that this is a very powerful feature but it also
2511 can easily lead to mistakes. If you really don't need this feature you
2512 can turn it off by putting the line:
2516 at the beginning of the phonetic table file. As mentioned, for rules
2517 containing a `-' no follow-up rules are searched but giving such rules
2518 a priority is not totally senseless because they can be follow-up rules
2519 and in that case the priority makes sense again. Follow-up rules of
2520 follow-up rules are not searched because this is in fact not needed
2523 The control character `^' says that the search string only matches
2524 at the beginning of words so that the rule `RH -> R' will only apply to
2525 words like `RHESUS' but not `PERHAPS'. You can append another `^' to
2526 the search string. In that case the algorithm treats the rest of the
2527 word totally separately from the first matched string at the beginning.
2528 This is useful for prefixes whose pronunciation does not depend on the
2529 rest of the word and vice versa like `OVER^^' in English for example.
2531 The same way as `^' works does `$' only apply to words that end with
2532 the search string. `GN$ -> N' only matches on words like `SIGN' but
2533 not `SIGNUM'. If you use `^' and `$' together, both of them must fit
2534 `ENOUGH^$ -> NF' will only match the word `ENOUGH' and nothing else.
2536 Of course you can combine all of the mentioned control characters but
2537 they must occur in this order: `< - priority ^ $'. All characters must
2538 be written in CAPITAL letters.
2540 If absolutely no rule can be found -- might happen if you use strange
2541 characters for which you don't have any replacement rule -- the next
2542 character will simply be skipped and the search for replacement rules
2543 will continue with the rest of the word.
2545 If you want double letters to be reduced to one you must set up a
2546 rule like `LL- -> L'. If double letters in the resulting phonetic word
2547 should be allowed, you must place the line:
2551 at the beginning of your transformation table file; otherwise set the
2552 value to `1'. The English rules for example strip all vowels from
2553 words and so the word "GOGO" would be transformed to "K" and not to
2554 "KK" (as desired) if `collapse_result' is set to 1. That's why the
2555 English rules have `collapse_result' set to `0'.
2557 By default, all accents are removed from a word before it is matched
2558 to the soundslike rules. If you do not want this then add the line
2562 at the beginning of your file. The exact definition of an accent is
2563 language dependent and is controlled via the character set file. If you
2564 set remove_accents to '0' then you should also set "store-as" to "lower"
2565 in the language data file (not the phonetic transformation file)
2566 otherwise Aspell will have problems when both the accented and the
2567 de-accented version of a word appearing in the dictionary; it will
2568 consider one of them as incorrectly spelled.
2570 7.3.2 How do I start finally?
2571 -----------------------------
2573 Before you start to write an array of transformation rules, you should
2574 be aware that you have to do some work to make sure that things you do
2575 will result in correct transformation rules.
2577 7.3.2.1 Things that come in handy
2578 .................................
2580 First of all, you need to have a large word list of the language you
2581 want to make phonetics for. It should contain about as many words as
2582 the dictionary of the spell checker. If you don't have such a list,
2583 you will probably find an Ispell dictionary at
2584 `http://fmg-www.cs.ucla.edu/geoff/ispell-dictionaries.html' which will
2585 help you. You can then make affix expansion via `ispell -e' and then
2586 pipe it through `tr " " "\n"' to put one word on each line. After that
2587 you eventually have to convert special characters like `é' from
2588 Ispell's internal representation to latin1 encoding. `sed s/e'/é/g'
2589 for example would replace all `e'' with `é'.
2591 The second is that you know how to use regular expressions and know
2592 how to use `grep'. You should for example know that:
2594 grep ^[^aeiou]qu[io] wordlist | less
2596 will show you all words that begin with any character but `a', `e',
2597 `i', `o' or `u' and then continue with `qui' or `quo'. This stuff is
2598 important for example to find out if a phonetic replacement rule you
2599 want to set up is valid for all words which match the expression you
2600 want to replace. Taking a look at the regex(7) man page is a good idea.
2602 7.3.2.2 What the phonetic code should do
2603 ........................................
2605 Normal text comparison works well as long as the typer misspells a word
2606 because he pressed one key he didn't really want to press. In these
2607 cases, mostly one character differs from the original word.
2609 In cases where the writer didn't know about the correct spelling of
2610 the word, the word may have several characters that differ from the
2611 original word but usually the word would still sound like the original.
2612 Someone might think that `tough' is spelled `taff'. No spell checker
2613 without phonetic code will come to the idea that this might be `tough',
2614 but a spell checker who knows that `taff' would be pronounced like
2615 `tough' will make good suggestions to the user. Another example could
2616 be `funetik' and `phonetic'.
2618 From these examples you can see that the phonetic transformation
2619 should not be too fussy and too precise. If you implement a whole
2620 phonetic dictionary as you can find it in books this will not be very
2621 useful because then there could still be many characters differing from
2622 the misspelled and the desired word. What you should do if you
2623 implement the phonetic transformation table is to reduce the number of
2624 used letters to the only really necessary ones.
2626 Characters that sound similar should be reduced to one. In the
2627 English language for example `Z' sounds like `S' and that's why the
2628 transformation rule `Z -> S' is present in the replacement table. "PH
2629 is spoken like "F and so we have a `PH -> F' rule.
2631 If you take a closer look you will even see that vowels sound very
2632 similar in the English language: `contradiction', `cuntradiction',
2633 `cantradiction' or `centradiction' in fact sound nearly the same, don't
2634 they? Therefore the English phonetic replacement rules not only reduce
2635 all vowels to one but even remove them all (removing is done by just
2636 setting up no rule for those letters). The phonetic code of
2637 "contradiction" is "KNTRTKXN" and if you try to read this
2638 letter-monster loud you will hear that it still sound a bit like
2639 `contradiction'. You also see that "D" is transformed to "T" because
2640 they nearly sound the same.
2642 If you think you have found a regularity you should _always_ take
2643 your word list and `grep' for the corresponding regular expression you
2644 want to make a transformation rule for. An example: If you come to the
2645 idea that all English words ending on `ough' sound like `AF' at the end
2646 because you think of `enough' and `tough'. If you then `grep' for the
2647 corresponding regular expression by `grep -i ough$ wordlist' you will
2648 see that the rule you wanted to set up is not correct because the rule
2649 doesn't fit to words like `although' or `bough'. So you have to define
2650 your rule more precisely or you have to set up exceptions if the number
2651 of words that differ from the desired rule is not too big.
2653 Don't forget about follow-up rules which can help in many cases but
2654 which also can lead to confusion and unwanted side effects. It's also
2655 important to write exceptions in front of the more general rules (`GH'
2658 If you think you have set up a number of rules that may produce some
2659 good results try them out! If you run Aspell as `aspell
2660 --lang=YOUR_LANGUAGE pipe' you get a prompt at which you can type in
2661 words. If you just type words Aspell checks them and eventually makes
2662 suggestions if they are misspelled. If you type in `$$Sw WORD' you
2663 will see the phonetic transformation and you can test out if your work
2666 Another good way to check that changes you make to your rules don't
2667 have any bad side effects is to create another list from your word list
2668 which contains not only the word of the word list but also the
2669 corresponding phonetic version of this word on the same line. If you
2670 do this once before the change and once after the change you can make a
2671 diff (see `man diff') to see what _really_ changed. To do this use the
2672 command `aspell --lang=YOUR_LANGUAGE soundslike'. In this mode Aspell
2673 will output the the original word and then its soundslike separated by
2674 a tab character for each word you give it. If you are interested in
2675 seeing how the algorithm works you can download a set of useful
2677 `http://members.xoom.com/maccy/spell/phonet-utils.tar.gz'. This
2678 includes a program that produces a list as mentioned above and another
2679 program which illustrates how the algorithm works. It uses the same
2680 transformation table as Aspell and so it helps a lot during the process
2681 of creating a phonetic transformation table for Aspell.
2683 During your work you should write down your basic ideas so that other
2684 people are able to understand what you did (and you still know about it
2685 after a few weeks). The English table has a huge documentation
2686 appended as an example.
2688 Now you can start experimenting with all the things you just read and
2689 perhaps set up a nice phonetic transformation table for your language
2690 to help Aspell to come up with the best correction suggestions ever
2691 seen also for your language. Take a look at the Aspell homepage to see
2692 if there is already a transformation table for your language. If there
2693 is one you might also take a look at it to see if it could be improved.
2695 If you think that this section helped you or if you think that this
2696 is just a waste of time you can send any feedback to
2697 <bjoern.jacke@gmx.de>.
2700 File: aspell.info, Node: The Simple Soundslike, Next: Replacement Tables, Prev: Phonetic Code, Up: Adding Support For Other Languages
2702 7.4 The Simple Soundslike
2703 =========================
2705 The simple soundslike goes something like this:
2707 sl0[0] = lookup0(word[0])
2708 for (i = 1; i < size; i++)
2709 sl0[i] = lookup(word[i]);
2711 for (i = 0; i < size; i++)
2712 sl.append(al0[i]) unless sl0[i] == 0 || sl0[i] == sl0[i-1];
2714 Basically each character can be converted to another character or
2715 deleted. A separate lookup table is used for the first character. If
2716 the same soundslike letter is repeated, the duplicate is removed.
2718 By default all accents are removed, and all vowels are deleted unless
2719 they appear at the start of the word in which case they are converted
2720 to a `*'. The exact behavior can be customized via the character data
2723 The simplified soundslike has the advantage that it is very fast to
2724 compute and thus does not need to be stored with a word. Also, when
2725 affix compression is used and the `partially-expand' is given the
2726 results will be identical to the results when affix compression is not
2729 Of course it is not nearly as powerful as the phonetic soundslike.
2732 File: aspell.info, Node: Replacement Tables, Next: Affix Compression, Prev: The Simple Soundslike, Up: Adding Support For Other Languages
2734 7.5 Replacement Tables
2735 ======================
2737 When phonetic code is not used a replacement table can be used instead.
2738 To enable the use of a replacement table add the line `repl-table
2739 LANG', in which case the replacement table is excepted to be in the
2740 file `LANG_repl.dat'. A complete file name can also be specified in
2741 place of LANG. For compatibility with MySpell the replacement table
2742 can also be part of the affix file, in which case `repl-table' will be
2745 Replacement table syntax:
2747 REP [number_of_replacement_definitions]
2748 REP [what] [replacement]
2749 REP [what] [replacement]
2751 For example a possible English replacement table definition to
2752 handle misspelled consonants:
2765 File: aspell.info, Node: Affix Compression, Next: Controlling the Behavior of Run-together Words, Prev: Replacement Tables, Up: Adding Support For Other Languages
2767 7.6 Affix Compression
2768 =====================
2770 Aspell, as of version 0.60, now has support for affix compression. The
2771 codebase comes from MySpell found in OpenOffice.
2773 To add support for affix compression add the following lines to the
2779 The line `affix LANG' adds support for recognizing affix
2780 information, and the line `affix-compress true' enables affix
2783 The affix file is expected to be named `LANG_affix.dat'. It is the
2784 exact same format as those used by MySpell. More information can be
2785 found in the myspell/ directory of the distribution or at
2786 `http://lingucomponent.openoffice.org/dictionary.html'.
2788 Affix compression can also be used with soundslike lookup. Aspell
2789 does this by only storing the soundslike for the root word. When a
2790 word is misspelled it will search for a soundslike close to all
2791 possible roots of the misspelled word.
2793 When no soundslike information, or the simple soundslike, is used it
2794 may be beneficial to specify the option `partially-expand' which will
2795 partially expand a word with affix information so that the affix flags
2796 do not affect the first 3 letters of the word. This will allow Aspell
2797 to get more accurate results when scanning the list for near misses
2798 since the full word can be used and not just the root. Specifying this
2799 option, however, will also effectively expand any prefixes. Thus this
2800 option should not be used for prefix heavy languages such as Hebrew.
2802 An existing word list, without affix info, can be affix compressed
2803 using using `aspell munch-list'.
2805 7.6.1 Format of the Affix File
2806 ------------------------------
2808 An affix is either a prefix or a suffix attached to root words to make
2809 other words. For example supply -> supplied by dropping the "y" and
2810 adding an "ied" (the suffix).
2812 Here is an example of how to define one specific suffix borrowed
2813 from the English affix file.
2817 SFX D y ied [^aeiou]y
2821 This file is space delimited and case sensitive. So this information
2822 can be interpreted as follows:
2824 The first line has 4 fields:
2826 1 SFX indicates this is a suffix
2827 2 D is the name of the character which represents this suffix
2828 3 Y indicates it can be combined with prefixes (cross product)
2829 4 4 indicates that sequence of 4 affix entries are needed to
2830 properly store the affix information
2832 The remaining lines describe the unique information for the 4 affix
2833 entries that make up this affix. Each line can be interpreted as
2834 follows: (note fields 1 and 2 are used as a check against line 1 info)
2836 1 SFX indicates this is a suffix
2837 2 D is the name of the character which represents this affix
2838 3 y the string of chars to strip off before adding affix (a 0
2839 here indicates the NULL string)
2840 4 ied the string of affix characters to add (a 0 here indicates
2842 5 [^aeiou]y the conditions which must be met before the affix can be
2845 Field 5 is interesting. Since this is a suffix, field 5 tells us
2846 that there are 2 conditions that must be met. The first condition is
2847 that the next to the last character in the word must _not_ be any of the
2848 following "a", "e", "i", "o" or "u". The second condition is that the
2849 last character of the word must end in "y".
2851 7.6.2 When Compared With Ispell
2852 -------------------------------
2854 Now for comparison purposes, here is the same information from the
2855 Ispell `english.aff' compression file which was used as the basis for
2859 E > D # As in create > created
2860 [^AEIOU]Y > -Y,IED # As in imply > implied
2861 [^EY] > ED # As in cross > crossed
2862 [AEIOU]Y > ED # As in convey > conveyed
2864 The Ispell information has exactly the same information but in a
2865 slightly different (case-insensitive) format:
2867 Here are the ways to see the mapping from Ispell .aff format to our
2870 1. The Ispell english.aff has flag D under the "suffix" section so
2871 you know it is a suffix.
2873 2. The D is the character assigned to this suffix
2875 3. `*' indicates that it can be combined with prefixes
2877 4. Each line following the : describes the affix entries needed to
2880 * The first field is the conditions that must be met.
2882 * The second field is after the > if a "-" occurs is the string
2883 to strip off (can be blank).
2885 * The third field is the string to add (the affix)
2887 In addition all chars in Ispell aff files are in uppercase.
2889 7.6.3 Specifying Affix Flags
2890 ----------------------------
2892 Affix flags are specified in the word list by specifying them after the
2901 will associate the `D' and `G' flag with the word create.
2904 File: aspell.info, Node: Controlling the Behavior of Run-together Words, Next: Creating A New Character Set, Prev: Affix Compression, Up: Adding Support For Other Languages
2906 7.7 Controlling the Behavior of Run-together Words
2907 ==================================================
2909 Aspell currently has support for unconditionally accepting run-together
2912 Support for unconditionally accepting run-together words can either
2913 be turned on in the language data file or as a normal option via the
2914 `run-together' option. The `run-together-limit' options controls the
2915 maximum number of words that can be strung together, the default is
2916 normally 2. The `run-together-min' options controls the minimum length
2917 of the individual components of the run together word, the default is
2918 normally 3. Both the `run-together-limit' and `run-together-min'
2919 option may be specified in both the language data file or as a normal
2923 File: aspell.info, Node: Creating A New Character Set, Next: Creating An Official Dictionary Package, Prev: Controlling the Behavior of Run-together Words, Up: Adding Support For Other Languages
2925 7.8 Creating A New Character Set
2926 ================================
2928 If there is not a standard character set for your language then you can
2929 invent one. The new charset will only be used by Aspell internally.
2930 If the option `data-encoding' is set to `utf-8', and your current
2931 locale character type is always set to `utf-8', then you can use UTF-8
2932 for everything and not worry yourself that an 8-bit character set is
2933 being used internally. If your language has no more than 210 distinct
2934 symbols, including different capitalizations and accents, then Aspell
2937 The first thing to do is to download the Aspell lang package (*note
2938 Creating An Official Dictionary Package::) and check if one of the
2939 provided charsets in this package will suite your needs. Non-standard
2940 character sets are provided for many scripts and languages. If not,
2941 then see the included `README' file for instructions on creating a new
2942 one. Version 0.1, and 0.2 of mkchardata _will not_ work as the format
2943 of the character data file has changed.
2946 File: aspell.info, Node: Creating An Official Dictionary Package, Prev: Creating A New Character Set, Up: Adding Support For Other Languages
2948 7.9 Creating An Official Dictionary Package
2949 ===========================================
2951 Once you have a basic dictionary working, you should consider creating
2952 an official package so that it can be distributed with Aspell. To do so
2953 download the aspell-lang package available at
2954 `ftp://ftp.gnu.org/gnu/aspell/aspell-lang-VERSION.tar.bz2' or in the
2955 "aspell-lang" module in the Aspell CVS repository available at
2956 `https://savannah.gnu.org/cvs/?group=aspell'. See the included
2957 `README' file for what to do. Or, send mail to aspell-dict at gnu org
2958 asking for help on how to get started.
2961 File: aspell.info, Node: Implementation Notes, Next: Languages Which Aspell can Support, Prev: Adding Support For Other Languages, Up: Top
2963 Appendix A Implementation Notes
2964 *******************************
2968 * Aspell Suggestion Strategy::
2969 * Notes on 8-bit Characters::
2972 File: aspell.info, Node: Aspell Suggestion Strategy, Next: Notes on 8-bit Characters, Up: Implementation Notes
2974 A.1 Aspell Suggestion Strategy
2975 ==============================
2977 The magic behind my spell checker comes from merging Lawrence Philips
2978 excellent metaphone algorithm and Ispell's near miss strategy which is
2979 inserting a space or hyphen, interchanging two adjacent letters,
2980 changing one letter, deleting a letter, or adding a letter.
2982 The process goes something like this.
2984 1. Convert the misspelled word to its soundslike equivalent (its
2985 metaphone for English words).
2987 2. Find all words that have a soundslike within one or two edit
2988 distances from the original word's soundslike. The edit distance
2989 is the total number of deletions, insertions, exchanges, or
2990 adjacent swaps needed to make one string equivalent to the other.
2991 When set to only look for soundslikes within one edit distance it
2992 tries all possible soundslike combinations and checks if each one
2993 is in the dictionary. When set to find all soundslike within two
2994 edit distances it scans through the entire dictionary and quickly
2995 scores each soundslike. The scoring is quick because it will give
2996 up if the two soundslikes are more than two edit distances apart.
2998 3. Find misspelled words that have a correctly spelled replacement by
2999 the same criteria of step number 2 and 3. That is the misspelled
3000 word in the word pair (such as "teh -> the") would appear in the
3001 suggestions list as if it was a correct spelling.
3003 4. Score the result list and return the words with the lowest score.
3004 The score is roughly the weighed average of the weighed edit
3005 distance of the word to the misspelled word and the soundslike
3006 equivalent of the two words. The weighted edit distance is like
3007 the edit distance except that the various edits have weights
3010 5. Replace the misspelled words that have correctly spelled
3011 replacements with their replacements and remove any duplicates
3012 that might arise because of this.
3014 Please note that the soundslike equivalent is a rough approximation
3015 of how the words sounds. It is not the phoneme of the word by any
3016 means. For more details about exactly how each step is performed
3017 please see the file `suggest.cc'. For more information on the metaphone
3018 algorithm please see the data file `english_phonet.dat'.
3021 File: aspell.info, Node: Notes on 8-bit Characters, Prev: Aspell Suggestion Strategy, Up: Implementation Notes
3023 A.2 Notes on 8-bit Characters
3024 =============================
3026 There is a very good reason I use 8-bit characters in Aspell. Speed and
3027 simplicity. While many parts of my code can fairly easily be converted
3028 to some sort of wide character as my code is clean. Other parts cannot
3031 One of the reasons why is because in many, many places I use a direct
3032 lookup to find out various information about characters. With 8-bit
3033 characters this is very feasible because there is only 256 of them.
3034 With 16-bit wide characters this will waste a LOT of space. With 32-bit
3035 characters this is just plain impossible. Converting the lookup tables
3036 to another form is certainly possible but degrades performance
3039 Furthermore, some of my algorithms rely on words consisting only on a
3040 small number of distinct characters (often around 30 when case and
3041 accents are not considered). When the possible character can consist of
3042 any Unicode character this number becomes several thousand, if that. In
3043 order for these algorithms to still be used, some sort of limit will
3044 need to be placed on the possible characters the word can contain. If I
3045 impose that limit, I might as well use some sort of 8-bit characters
3046 set which will automatically place the limit on what the characters can
3049 There is also the issue of how I should store the word lists in
3050 memory? As a string of 32 bit wide characters. Now that is using up 4
3051 times more memory than characters would and for languages that can fit
3052 within an 8-bit character that is, in my view, a gross waste of memory.
3053 So maybe I should store them is some variable width format such as
3054 UTF-8. Unfortunately, way, way too many of the algorithms will simply
3055 not work with variable width characters without significant
3056 modification which will very likely degrade performance. So the
3057 solution is to work with the characters as 32-bit wide characters and
3058 then convert it to a shorter representation when storing them in the
3059 lookup tables. Now that can lead to an inefficiency. I could also use
3060 16 bit wide characters, however that may not be good enough to hold all
3061 future versions of Unicode and therefore has the same problems.
3063 As a response to the space waste used by storing word lists in some
3064 sort of wide format some one asked:
3066 Since hard drives are cheaper and cheaper, you could store a
3067 dictionary in a usable (uncompressed) form and use it directly
3068 with memory mapping. Then the efficiency would directly depend on
3069 the disk caching method, and only the used part of the
3070 dictionaries would really be loaded into memory. You would no more
3071 have to load plain dictionaries into main memory, you'll just want
3072 to compute some indexes (or something like that) after mapping.
3074 However, the fact of the matter is that most of the dictionary will
3075 be read into memory anyway if it is available. If it is not available
3076 then there would be a good deal of disk swaps. Making characters 32-bit
3077 wide will increase the chance that there are more disk swaps. So the
3078 bottom line is that it is more efficient to convert characters from
3079 something like UTF-8 into some sort of 8-bit character. I could also
3080 use some sort of disk space lookup table such as the Berkeley Database.
3081 However this will *definitely* degrade performance.
3083 The bottom line is that keeping Aspell 8-bit internally is a very
3084 well though out decision that is not likely to change any time soon.
3085 Feel free to challenge me on it, but, don't expect me to change my mind
3086 unless you can bring up some point that I have not thought of before
3087 and quite possibly a patch to solve cleanly convert Aspell to Unicode
3088 internally without a serious performance lost OR serious memory usage
3092 File: aspell.info, Node: Languages Which Aspell can Support, Next: Language Related Issues, Prev: Implementation Notes, Up: Top
3094 Appendix B Languages Which Aspell can Support
3095 *********************************************
3097 Even though Aspell will remain 8-bit internally it should still be able
3098 to support any written languages not based on a logographic script.
3099 The only logographic writing system in current use are those based on
3100 hànzi which includes Chinese, Japanese, and sometimes Korean.
3106 * Multiple Scripts::
3107 * Planned Dictionaries::
3111 File: aspell.info, Node: Supported, Next: Unsupported, Up: Languages Which Aspell can Support
3116 Aspell 0.60 should be able to support the following languages:
3118 Code Language Name Script Dictionary Gettext
3119 Available Translation
3122 af Afrikaans Latin 0.50 -
3123 ak Akan Latin Maybe -
3124 am Amharic Ethiopic 0.60 -
3125 ar Arabic Arabic 0.60 -
3126 as Assamese Bengali - -
3127 av Avar Cyrillic - -
3129 az Azerbaijani Cyrillic, Latin 0.60 -
3131 ba Bashkir Cyrillic - -
3132 be Belarusian Cyrillic 0.50 Incomplete
3133 bg Bulgarian Cyrillic 0.50 -
3134 bh Bihari Devanagari - -
3135 bm Bambara Latin - -
3136 bn Bengali Bengali 0.60 -
3137 bo Tibetan Tibetan - -
3138 br Breton Latin 0.50 -
3139 bs Bosnian Latin Maybe -
3141 ca Catalan / Valencian Latin 0.50 Yes
3142 ce Chechen Cyrillic - -
3143 co Corsican Latin Maybe -
3144 cop Coptic Greek Maybe -
3145 cs Czech Latin 0.50 Yes
3146 csb Kashubian Latin 0.60 -
3147 cv Chuvash Cyrillic - -
3148 cy Welsh Latin 0.50 -
3150 da Danish Latin 0.50 Incomplete
3151 de German Latin 0.50 Yes
3155 el Greek Greek 0.50 -
3156 en English Latin 0.50 Yes
3157 eo Esperanto Latin 0.50 -
3158 es Spanish Latin 0.50 Incomplete
3159 et Estonian Latin 0.60 -
3160 eu Basque Latin Maybe -
3162 fa Persian Arabic 0.60 -
3163 ff Fulah Latin Maybe -
3164 fi Finnish Latin 0.60 -
3165 fj Fijian Latin Maybe -
3166 fo Faroese Latin 0.50 -
3167 fr French Latin 0.50 Yes
3168 fur Friulian Latin Maybe -
3169 fy Frisian Latin 0.60 -
3171 ga Irish Latin 0.50 Yes
3172 gd Scottish Gaelic Latin 0.50 -
3173 gl Gallegan Latin 0.50 -
3174 gn Guarani Latin Maybe -
3175 gu Gujarati Gujarati 0.60 -
3176 gv Manx Gaelic Latin 0.50 -
3178 ha Hausa Latin Maybe -
3179 he Hebrew Hebrew 0.60 -
3180 hi Hindi Devanagari 0.60 -
3181 hil Hiligaynon Latin 0.50 -
3182 ho Hiri Motu Latin - -
3183 hr Croatian Latin 0.50 -
3184 hsb Upper Sorbian Latin 0.60 -
3185 ht Haitian Creole Latin Maybe -
3186 hu Hungarian Latin 0.60 -
3187 hy Armenian Armenian 0.60 -
3190 ia Interlingua (IALA) Latin 0.50 -
3191 id Indonesian Arabic, Latin 0.50 -
3192 ig Igbo Latin Maybe -
3193 ii Sichuan Yi Yi - -
3195 is Icelandic Latin 0.50 -
3196 it Italian Latin 0.50 Yes
3198 jv Javanese Javanese, Latin Maybe -
3200 ka Georgian Georgian - -
3201 kg Kongo Latin Maybe -
3202 ki Kikuyu / Gikuyu Latin - -
3203 kj Kwanyama Latin - -
3204 kk Kazakh Cyrillic - -
3205 km Khmer Khmer Maybe -
3206 kn Kannada Kannada Planned -
3208 ks Kashmiri Arabic, Devanagari - -
3209 ku Kurdish Arabic, Cyrillic, 0.50 -
3211 kv Komi Cyrillic - -
3212 ky Kirghiz Arabic, Cyrillic, Maybe -
3215 la Latin Latin 0.60 -
3216 lb Luxembourgish Latin Maybe -
3217 lg Ganda Latin Maybe -
3218 li Limburgian Latin Maybe -
3219 ln Lingala Latin Maybe -
3220 lt Lithuanian Latin 0.60 -
3221 lu Luba-Katanga Latin - -
3222 lv Latvian Latin 0.60 -
3224 mg Malagasy Latin 0.50 -
3225 mi Maori Latin 0.50 -
3226 mk Macedonian Cyrillic 0.50 -
3227 ml Malayalam Latin, Malayalam 0.60 -
3228 mn Mongolian Cyrillic, Mongolian 0.60 Incomplete
3229 mo Moldavian Cyrillic - -
3231 mr Marathi Devanagari 0.60 -
3232 ms Malay Arabic, Latin 0.50 -
3233 mt Maltese Latin 0.50 -
3234 my Burmese Myanmar - -
3236 nb Norwegian Bokmal Latin 0.50 -
3237 nd North Ndebele Latin Maybe -
3238 nds Low Saxon Latin 0.60 -
3239 ne Nepali Devanagari Maybe -
3240 ng Ndonga Latin Maybe -
3241 nl Dutch Latin 0.50 Yes
3242 nn Norwegian Nynorsk Latin 0.50 -
3243 nr South Ndebele Latin Maybe -
3244 nso Northern Sotho Latin Maybe -
3245 nv Navajo Latin Maybe -
3246 ny Nyanja Latin 0.50 -
3248 oc Occitan / Provencal Latin Maybe -
3249 om Oromo Ethiopic, Latin - -
3250 or Oriya Oriya 0.60 -
3251 os Ossetic Cyrillic - -
3253 pa Punjabi Gurmukhi 0.60 -
3254 pl Polish Latin 0.50 -
3255 ps Pushto Arabic - -
3256 pt Portuguese Latin 0.50 Incomplete
3258 qu Quechua Latin 0.60 -
3260 rn Rundi Latin Maybe -
3261 ro Romanian Latin 0.50 Incomplete
3262 ru Russian Cyrillic 0.50 Yes
3263 rw Kinyarwanda Latin 0.50 -
3265 sc Sardinian Latin 0.50 -
3266 sd Sindhi Arabic - -
3267 sg Sango Latin Maybe -
3268 si Sinhalese Sinhala - -
3269 sk Slovak Latin 0.50 Yes
3270 sl Slovenian Latin 0.50 Yes
3271 sm Samoan Latin Maybe -
3272 sn Shona Latin Maybe -
3273 so Somali Latin Maybe -
3274 sq Albanian Latin Maybe -
3275 sr Serbian Cyrillic, Latin 0.60 Incomplete
3276 ss Swati Latin Maybe -
3277 st Southern Sotho Latin Maybe -
3278 su Sundanese Latin Maybe -
3279 sv Swedish Latin 0.50 Incomplete
3280 sw Swahili Latin 0.50 -
3282 ta Tamil Tamil 0.60 -
3283 te Telugu Telugu 0.60 -
3284 tet Tetum Latin 0.50 -
3285 tg Tajik Arabic, Cyrillic, Maybe Incomplete
3287 ti Tigrinya Ethiopic Maybe -
3288 tk Turkmen Arabic, Cyrillic, 0.50 -
3290 tl Tagalog Latin, Tagalog 0.50 -
3291 tn Tswana Latin 0.50 -
3292 to Tonga Latin Maybe -
3293 tr Turkish Arabic, Latin 0.50 -
3294 ts Tsonga Latin Maybe -
3295 tt Tatar Cyrillic - -
3297 ty Tahitian Latin Maybe -
3299 ug Uighur Arabic, Cyrillic, - -
3301 uk Ukrainian Cyrillic 0.50 Yes
3302 ur Urdu Arabic Maybe -
3303 uz Uzbek Cyrillic, Latin 0.60 -
3305 ve Venda Latin Maybe -
3306 vi Vietnamese Latin 0.60 Yes
3308 wa Walloon Latin 0.50 Incomplete
3309 wo Wolof Latin Maybe -
3311 xh Xhosa Latin Maybe -
3313 yi Yiddish Hebrew 0.60 -
3314 yo Yoruba Latin Maybe -
3317 zu Zulu Latin 0.50 -
3319 Dictionaries marked as "0.50" are available for Aspell 0.50. Ones
3320 marked as "0.60" are available for Aspell 0.60 only. Ones marked as
3321 "Planned" should eventually be available. Ones marked as "Maybe" might
3322 be available in the future. *Note Planned Dictionaries::, for more
3325 B.1.1 Notes on Latin Languages
3326 ------------------------------
3328 Any word that can be written using one of the Latin ISO-8859 character
3329 sets (ISO-8859-1,2,3,4,9,10,13,14,15,16) can be written, in decomposed
3330 form, using the ASCII characters, the 23 additional letters:
3332 U+00C6 LATIN CAPITAL LETTER AE
3333 U+00D0 LATIN CAPITAL LETTER ETH
3334 U+00D8 LATIN CAPITAL LETTER O WITH STROKE
3335 U+00DE LATIN CAPITAL LETTER THORN
3336 U+00DE LATIN SMALL LETTER THORN
3337 U+00DF LATIN SMALL LETTER SHARP S
3338 U+00E6 LATIN SMALL LETTER AE
3339 U+00F0 LATIN SMALL LETTER ETH
3340 U+00F8 LATIN SMALL LETTER O WITH STROKE
3341 U+0110 LATIN CAPITAL LETTER D WITH STROKE
3342 U+0111 LATIN SMALL LETTER D WITH STROKE
3343 U+0126 LATIN CAPITAL LETTER H WITH STROKE
3344 U+0127 LATIN SMALL LETTER H WITH STROKE
3345 U+0131 LATIN SMALL LETTER DOTLESS I
3346 U+0138 LATIN SMALL LETTER KRA
3347 U+0141 LATIN CAPITAL LETTER L WITH STROKE
3348 U+0142 LATIN SMALL LETTER L WITH STROKE
3349 U+014A LATIN CAPITAL LETTER ENG
3350 U+014B LATIN SMALL LETTER ENG
3351 U+0152 LATIN CAPITAL LIGATURE OE
3352 U+0153 LATIN SMALL LIGATURE OE
3353 U+0166 LATIN CAPITAL LETTER T WITH STROKE
3354 U+0167 LATIN SMALL LETTER T WITH STROKE
3356 and the 14 modifiers:
3358 U+0300 COMBINING GRAVE ACCENT
3359 U+0301 COMBINING ACUTE ACCENT
3360 U+0302 COMBINING CIRCUMFLEX ACCENT
3361 U+0303 COMBINING TILDE
3362 U+0304 COMBINING MACRON
3363 U+0306 COMBINING BREVE
3364 U+0307 COMBINING DOT ABOVE
3365 U+0308 COMBINING DIAERESIS
3366 U+030A COMBINING RING ABOVE
3367 U+030B COMBINING DOUBLE ACUTE ACCENT
3368 U+030C COMBINING CARON
3369 U+0326 COMBINING COMMA BELOW
3370 U+0327 COMBINING CEDILLA
3371 U+0328 COMBINING OGONEK
3373 Which is a total of 37 additional Unicode code points.
3375 All ISO-8859 character leaves the characters 0x00 - 0x1F, and 0x80 -
3376 0x9F unmapped as they are generally used as control characters. Of
3377 those, 0x01 - 0x0F, 0x11 - 0x1F and 0x80 - 0x9F may be mapped to
3378 anything in Aspell. This is a total of 62 characters which can be
3379 remapped in any ISO-8859 character set. Thus, by remapping 37 of the 62
3380 characters to the previously specified Unicode code-points, any modified
3381 ISO-8859 character set can be used for any Latin languages covered by
3382 ISO-8859. Of course decomposing every single accented character wastes
3383 a lot of space, so only characters that cannot be represented in the
3384 precomposed form should be broken up. By using this trick it is
3385 possible to store foreign words in the correctly accented form in the
3386 dictionary even if the precomposed character is not in the current
3389 Any letter in the Unicode range U+0000 - U+0249, U+1E00 - U+1EFF
3390 (Basic Latin, Latin-1 Supplement, Latin Extended-A, Latin Extended-B,
3391 and Latin Extended Additional) can be represented using around 175
3392 basic letters, and 25 modifiers which is less than 210 and can thus fit
3393 in an Aspell 8-bit character set. Since this Unicode range covers any
3394 possible Latin language this special character set can be used to
3395 represent any word written using the Latin script if so desired.
3400 Syllabic languages use a separate symbol for each syllable of the
3401 language. Even thought most of them have more than 210 distinct
3402 symbols Aspell can still support them by breaking them up.
3404 B.1.2.1 The Ethiopic Syllabary
3405 ..............................
3407 Even though the Ethiopic script has more than 210 distinct characters
3408 Aspell can still handle it. The idea is to split each character into
3409 two parts based on the Consonant and Vowel parts. This encoding of the
3410 syllabary is far more useful to Aspell than if they were stored in UTF-8
3411 or UTF-16. In fact, the exiting suggestion strategy of Aspell will work
3412 well with this encoding without any additional modifications. However,
3413 additional improvements may be possible by taking advantage of the
3414 consonant-vowel structure of this encoding.
3416 In fact, the split consonant-vowel representation may prove to be so
3417 useful that it may be beneficial to encode other syllabary in this
3418 fashion, even if they are less than 210 of them.
3420 The code to break up a syllabary into the consonant-vowel part is
3421 part of the Unicode normalization process.
3423 B.1.2.2 The Yi Syllabary
3424 ........................
3426 A very large syllabary with 819 distinct symbols. However, like
3427 Ethiopic, it should be possible to support this script by breaking it
3430 B.1.2.3 The Ojibwe Syllabary
3431 ............................
3433 With only 120 distinct symbols, Aspell can actually support this one as
3434 is. However, as previously mentioned, it may be beneficial to break it
3435 up into the consonant-vowel representation anyway.
3438 File: aspell.info, Node: Unsupported, Next: Multiple Scripts, Prev: Supported, Up: Languages Which Aspell can Support
3443 These languages, when written in the given script, are currently
3444 unsupported by Aspell for one reason or another.
3446 Code Language Name Script
3447 ja Japanese Japanese
3449 ko Korean Han, Hangul
3454 B.2.1 The Thai, Khmer, and Lao Scripts
3455 --------------------------------------
3457 The Thai, Khmer, and Lao scripts presents a different problem for
3458 Aspell. The problem is not that there are more than 210 unique symbols,
3459 but that there are no spaces between words. This means that there is no
3460 easy way to split a sentence into individual words. However, it is
3461 still possible to spell check these scripts, it is just a lot more
3462 difficult. I will be happy to work with someone who is interested in
3463 adding Thai, Khmer, or Lao support to Aspell, but it is not likely
3464 something I will do on my own in the foreseeable future.
3466 B.2.2 Languages which use Hànzi Characters
3467 ------------------------------------------
3469 Hànzi Characters are used to write Chinese, Japanese, Korean, and were
3470 once used to write Vietnamese. Each hànzi character represents a
3471 syllable of a spoken word and also has a meaning. Since there are
3472 around 3,000 of them in common usage it is unlikely that Aspell will
3473 ever be able to support spell checking languages written using hànzi
3474 until full Unicode support is implemented. However, I am not even sure
3475 if these languages need spell checking since hànzi characters are
3476 generally not entered in directly. Furthermore even if Aspell could
3477 spell check hànzi the existing suggestion strategy will not work well
3478 at all, and thus a completely new strategy will need to be developed.
3479 However, if it is the case that hànzi needs to be spell checked and you
3480 know something about the issues involved please fell free to contact me.
3485 Modern Japanese is written in a mixture of "hiragana", "katakana",
3486 "kanji", and sometimes "romaji". "Hiragana" and "katakana" are both
3487 syllabaries unique to Japan, "kanji" is a modified form of hànzi, and
3488 "romaji" uses the Latin alphabet. With some work, Aspell should be
3489 able to check the non-kanji part of Japanese text. However, based on
3490 my limited understanding of Japanese hiragana is often used at the end
3491 of kanji. Thus if Aspell was to simply separate out the hiragana from
3492 kanji it would end up with a lot of word endings which are not proper
3493 words and will thus be flagged as misspellings. However, this can be
3494 fairly easily rectified as text is tokenized into words before it is
3495 converted into Aspell's internal encoding. In fact, some Japanese text
3496 is written in entirely in one script. For example books for children
3497 and foreigners are sometimes written entirely in hiragana. Thus,
3498 Aspell, in its current state, could prove at least somewhat useful for
3499 spell checking Japanese.
3504 Korean is generally written in hangul or a mixture of han and hangul.
3505 In Hangul letters individual letters, known as jamo, are grouped
3506 together in syllable blocks. Unicode allows Hangul to be stored in one
3507 of three ways, (A) Individual jamo letters (Hangul Compatibility Jamo,
3508 U+3130 - U+318F), (D) decomposed jamo (Hangul Jamo, U+1100 - U+11FF),
3509 and (C) precoposed sylable blocks (Hangul Syllables, U+AC00 - U+D7AF).
3510 In order for Aspell to work with Hangul it needs to be form A.
3511 Unfortunately the existing Normalization code in Aspell will not be
3512 able to adequately deal with converting Hangul from form D and C to
3513 form A and back again. However, once this code is written, Aspell
3514 should be able to spell check Hangul without any problem.
3517 File: aspell.info, Node: Multiple Scripts, Next: Planned Dictionaries, Prev: Unsupported, Up: Languages Which Aspell can Support
3519 B.3 Languages Written in Multiple Scripts
3520 =========================================
3522 Aspell should be able to check text written in the same language but in
3523 multiple scripts with some work. If the number of unique symbols in
3524 both scripts is less than 210, then a special character set can be used
3525 to allow both scripts to be encoded in the same dictionary. However
3526 this may not be the most efficient solution. An alternate solution is
3527 to store each script in its own dictionary and allow Aspell to choose
3528 the correct dictionary based on which script the given word is written
3529 in. Aspell currently does not support this mode of spell checking but
3530 it is something that I hope to eventually support.
3533 File: aspell.info, Node: Planned Dictionaries, Next: References, Prev: Multiple Scripts, Up: Languages Which Aspell can Support
3535 B.4 Notes on Planned Dictionaries
3536 =================================
3538 According to `http://wiki.services.openoffice.org/wiki/Dictionaries',
3539 Open Office dictionaries are available for the following languages, but
3540 no corresponding Aspell dictionary exists:
3554 * Luxembourgish (lb)
3560 * South Ndebele (nr)
3562 * Northern Sotho (nso)
3566 * Southern Sotho (st)
3574 If you are interested in converting any of them please coordinate your
3575 efforts with the dictionary author and submit it to aspell-dict at gnu
3576 org when you have something ready.
3578 An unofficial dictionary for Albanian (sq) is available at
3579 `http://psychology.rutgers.edu/~zaimi/software.html'. However, I can
3580 not find any contact information for the author, thus I have been
3581 unable to contact him. In addition an Albanian (sq) dictionary is
3582 available for Ispell at
3583 `http://www.7kosova.com/kde-shqip/ispell/ispell.html'. However, the
3584 raw word list is not provided and the author has not been responding to
3585 emails, possibly because he doesn't speak English. If you have any
3586 additional information on either of these dictionaries, or can speak
3587 Albanian and can translate for me please let me know at <kevina@gnu.org>
3589 An unofficial dictionary for Malayalam (ml) is available at
3590 `http://in.geocities.com/paivakil/downloads/aspell/'. I am working
3591 with the author to create an official one.
3593 Kevin Patrick Scannell has word lists available for the following
3594 languages based on his web crawling software
3595 (`http://borel.slu.edu/crubadan/') but needs someone to proofread them:
3599 * Asturian / Bable (ast)
3611 * Catalan / Valencian (ca)
3637 * Scottish Gaelic (gd)
3651 * Haitian Creole (ht)
3665 * Kalaallisut / Greenlandic (kl)
3673 * Luxembourgish (lb)
3683 * Luo (Kenya and Tanzania) (luo)
3697 * North Ndebele (nd)
3705 * Norwegian Nynorsk (nn)
3707 * Northern Sotho (nso)
3713 * Occitan / Provencal (oc)
3729 * Northern Sami (se)
3741 * Southern Sotho (st)
3779 If you are interested, please contact him at scannell at slu edu.
3781 A dictionary marked as "Planned" or "Maybe" but not listed in the
3782 section means that someone has expressed an interest in creating one.
3783 If you are interested in helping please contact me at <kevina@gnu.org>
3784 so that I can put you in touch with them.
3787 File: aspell.info, Node: References, Prev: Planned Dictionaries, Up: Languages Which Aspell can Support
3792 The information in this chapter was gathered from numerous sources,
3795 * ISO 639-2 Registration Authority,
3796 `http://www.loc.gov/standards/iso639-2/'
3798 * Languages and Scripts (Official Unicode Site),
3799 `http://www.unicode.org/onlinedat/languages-scripts.html'
3801 * Omniglot - a guide to written language, `http://www.omniglot.com/'
3803 * Wikipedia - The Free Encyclopedia, `http://wikipedia.org/'
3805 * Ethnologue - Languages of the World, `http://www.ethnologue.com/'
3807 * World Languages - The Ultimate Language Store,
3808 `http://www.worldlanguage.com/'
3810 * South African Languages Web, `http://www.languages.web.za/'
3812 * The Languages and Writing Systems of Africa (Global Advisor
3813 Newsletter), `http://www.intersolinc.com/newsletters/africa.htm'
3816 Special thanks goes to Era Eriksson for helping me with the
3817 information in this chapter.
3820 File: aspell.info, Node: Language Related Issues, Next: To Do, Prev: Languages Which Aspell can Support, Up: Top
3822 Appendix C Language Related Issues
3823 **********************************
3825 Here are some language related issues that a good spell checker needs to
3826 handle. If you have any more information about any of these issues, or
3827 of a new issue not discussed here, please email me at <kevina@gnu.org>.
3832 * Words With Symbols in Them::
3833 * Unicode Normalization::
3835 * Context Sensitive Spelling::
3838 File: aspell.info, Node: Compound Words, Next: Words With Symbols in Them, Up: Language Related Issues
3843 In some languages, such as German, it is acceptable to string two words
3844 together, thus forming a compound word. However, there are rules to
3845 when this can be done. Furthermore, it is not always sufficient to
3846 simply concatenate the two words. For example, sometimes a letter is
3847 inserted between the two words. Aspell currently has support for
3848 unconditionally stringing words together. I tried implementing more
3849 sophisticated support for compound words in Aspell but it was too
3850 limiting and no one used it.
3852 After receiving feedback from several people it seems that acceptable
3853 support for compound words involved two basically independent parts.
3854 If this is not sufficient for your language please let me know.
3859 Describes how the word needs to be changed when forming a compound
3861 CMP <flag> <strip> <add> <cond> <cond2>
3863 <flag> is the compound flag
3864 <strip> is the string to strip or 0 for the null string
3865 <add> is the string to add or 0 for the null string
3866 <cond> is the condition to match at the end of the current word
3867 <cond2> is the condition to match at the beginning of the next word
3869 All but the last field are the same as a suffix entry in the existing
3872 <cond> is a simplified regular expression. Some examples:
3879 It does not seem necessary to change the beginning of a word when
3885 Describes the position a word can appear in (beginning, middle, or end)
3886 and with which words.
3888 To do this each word can be assigned a category. Then each category
3889 can be given a set of rules to describe how it can be used in a
3890 compound word for example
3892 A + B: indicates that category A may appear at the beginning of a
3893 word when followed by a category B word. When combined it is then
3894 considered a category B word.
3895 A + C + B: here a C word may only appear between an A or B word
3901 I have not decided if a word should be allowed to belong to more than
3902 one category as a new category can be created in necessary to mean
3903 words in both category A and B for example.
3908 To implement support for compound words based on the above description
3909 the following will need to be done:
3911 1. expand the affix code to support special compound flags as
3912 described in part one
3914 2. write code to store the conditions as described in part two
3916 3. expand the compound checking code to check against the conditions
3918 4. expand the dictionary format to store the necessary compound info
3922 I don't know when I will be able to actually implement this. If you
3923 would like to try please let me know.
3926 File: aspell.info, Node: Words With Symbols in Them, Next: Unicode Normalization, Prev: Compound Words, Up: Language Related Issues
3928 C.2 Words With Spaces or Other Symbols in Them
3929 ==============================================
3931 Many languages, including English, have words with non-letter symbols in
3932 them. For example the apostrophe. These symbols generally appear in
3933 the middle of a word, but they can also appear at the end, such as in an
3934 abbreviation. If a symbol can _only_ appear as part of a word then
3935 Aspell can treat it as if it were a letter.
3937 However, the problem is most of these symbols have other uses. For
3938 example, the apostrophe is often used as a single quote and the
3939 abbreviations marker is also used as a period. Thus, Aspell cannot
3940 blindly treat them as if they were letters.
3942 Aspell currently handles the case where the symbol can only appear in
3943 the middle of the word fairly well. It simply assumes that if there is
3944 a letter both before and after the symbol than it is part of the word.
3945 This works most of the time but it is not fool proof. For example,
3946 suppose the user forgot to leave a space after the period:
3948 ... and the dog went up the tree.Then the cat ...
3950 Aspell would think "tree.Then" is one word. A better solution might be
3951 to then try to check "tree" and "Then" separately. But what if one of
3952 them is not in the dictionary? Should Aspell assume "tree.Then" is one
3955 The case where the symbol can appear at the beginning or end of the
3956 word is more difficult to deal with. The symbol may or may not
3957 actually be part of the word. Aspell currently handles this case by
3958 first trying to spell check the word with the symbol and if that fails,
3959 try it without. The problem is, if the word is misspelled, should
3960 Aspell assume the symbol belongs with the word or not? Currently
3961 Aspell assumes it does, which is not always the correct thing to do.
3963 Numbers in words present a different challenge to Aspell. If Aspell
3964 treats numbers as letters then every possible number a user might write
3965 in a document must be specified in the dictionary. This could easily
3966 be solved by having special code to assume all numbers are correctly
3967 spelled. Yet, what about something like "4th". Since the "th" suffix
3968 can appear after any number we are left with the same problem. The
3969 solution would be to have a special symbol for "any number".
3971 Words with spaces in them, such as foreign phrases, are even more
3972 trouble to deal with. The basic problem is that when tokenizing a
3973 string there is no good way to keep phrases together. One solution is to
3974 use trial and error. If a word is not in the dictionary try grouping it
3975 with the previous or next word and see if the combined word is in the
3976 dictionary. But what if the combined word is not, should the misspelled
3977 word be grouped when looking for suggestions? One solution is to also
3978 store each part of the phrase in the dictionary, but tag it as part of a
3979 phrase and not an independent word.
3981 To further complicate things, most applications that use spell
3982 checkers are accustom to parsing the document themselves and sending it
3983 to the spell checker a word at a time. In order to support words with
3984 spaces in them a more complicated interface will be required.
3987 File: aspell.info, Node: Unicode Normalization, Next: German Sharp S, Prev: Words With Symbols in Them, Up: Language Related Issues
3989 C.3 Unicode Normalization
3990 =========================
3992 Because Unicode contains a large number of precomposed characters there
3993 are multiple ways a character can be represented. For example letter ö
3994 can either be represented as
3996 U+00F6 LATIN SMALL LETTER O WITH DIAERESIS
3998 U+0061 LATIN SMALL LETTER O + U+0308 COMBINING DIAERESIS
4000 By performing normalization first, Aspell will only see one of these
4001 representations. The exact form of normalization depends on the
4002 language. Give the choice of:
4004 1. Precomposed character
4006 2. Base letter + combining character(s)
4010 if the precomposed character is in the target character set, then (1),
4011 if both base and combining character is present, then (2), otherwise
4014 Unicode Normalization is now implemented in Aspell 0.60.
4017 File: aspell.info, Node: German Sharp S, Next: Context Sensitive Spelling, Prev: Unicode Normalization, Up: Language Related Issues
4022 The German Sharp S or Eszett does not have an uppercase equivalent.
4023 Instead when `ß' is converted to `SS'. The conversion of `ß' to `SS'
4024 requires a special rule, and increases the length of a word, thus
4025 disallowing inplace case conversion. Furthermore, my general rule of
4026 converting all words to lowercase before looking them up in the
4027 dictionary won't work because the conversion of `SS' to lowercase is
4028 ambiguous; it can be `ss' or `ß'. I do plan on dealing with this
4032 File: aspell.info, Node: Context Sensitive Spelling, Prev: German Sharp S, Up: Language Related Issues
4034 C.5 Context Sensitive Spelling
4035 ==============================
4037 In some language, such as Luxembourgish, the spelling of a word depends
4038 on which words surround it. For example the the letter `n' at the end
4039 of a word will disappear if it is followed by another word starting
4040 with a certain letter such as an `s'. However, it can probably get
4041 more complicated than that. I would like to know how complicated before
4042 I attempt to implement support for context sensitive spelling.
4045 File: aspell.info, Node: To Do, Next: Installing, Prev: Language Related Issues, Up: Top
4054 * Notes on Various Items::
4057 File: aspell.info, Node: Important Items, Next: Other Items, Up: To Do
4062 Words in bold indicate how you should refer to the item when discussing
4063 it with me or others.
4065 D.1.1 Things that need to be done
4066 ---------------------------------
4068 These items need to be done before I consider Aspell finished. If you
4069 are interested in helping me with one of these tasks please email me.
4070 Good C++ skills are needed for most of these tasks involving coding.
4072 * Create a generic filter to handle multi-character letters such as
4073 `"a' or `\"a' for ä. This filter should make use of the already
4074 exiting normalization code if possible.
4076 * Make Aspell *Thread safe*. Even though Aspell itself is not
4077 multi-threaded I would like it to be thread safe so that it can be
4078 used by multi-threaded programs. There are several areas of Aspell
4079 that are potentially thread unsafe (such as accessing a global
4080 pool) and several classes which have the potential of being used
4081 by more than one thread (such as the personal dictionary). _[In
4084 * Enhance *ispell.el* so that it will work better with GNU Aspell.
4087 * Clean up copyright notices and bring the Aspell package up to *GNU
4088 Standards*. _[In Progress]_.
4090 D.1.2 Things I would like to get done
4091 -------------------------------------
4093 I would like to get these done. However, I may still consider Aspell
4094 finished without. They will probably eventually get implemented.
4095 However, I could still use help with them.
4097 * Better support for *compound words*. The support for _conditional_
4098 compound words found in Aspell versions 0.50 and earlier is no
4099 longer available since no one seems to be using it. Support for
4100 _unconditional_ compound words is still available. *Note Compound
4103 * Be able to accept *words with spaces in them* as many languages
4104 have words, such as a word in a foreign phrase, which only makes
4105 sense when followed by other words. *Note Words With Symbols in
4108 * Reorganize manual to make it easier to understand and to make it
4109 possible to break out useful man pages.
4111 * Support *soundslike lookup with affix compression*. I think it is
4112 possible, although I don't know how effective it will be. The
4113 basic idea is to affix compress the soundslike codes and then
4114 match the codes up with affix compressed words. If you are
4115 interested, email <aspell-devel@gnu.org>, and I will explain it in
4118 * Use Lawrence Philips' new *Double Metaphone algorithm*. See
4119 `http://aspell.net/metaphone/'. The main task involved here is
4120 converting the algorithm into table form. This will take some time
4121 but there is no real programming experience required. If you want
4122 to help with Aspell but don't have any real programming experience,
4123 this would be a great place to start.
4125 * Rank suggestions based on *frequency information*. Both global
4126 frequency and document specific frequency can be used. The latter
4127 will require that the whole document be made available to the
4128 spell checker. Also use frequency information to flag words which
4129 are found in the dictionary but not in common usage, and thus
4130 might not be what was intended.
4132 * Support a *"dual-script" mode* where Aspell can use a separate
4133 dictionary depending on which script it detects the current word
4134 in, the two dictionaries can have nothing in common, ie an English
4135 one and a Russian one for example. This will _not_ support two
4136 languages that use the same script as that is a lot more
4137 complicated. For example if the word is misspelled which
4138 dictionary should it use for the suggestions?
4140 * Write a *GUI* for the Aspell utility. Ideally it should be able to
4141 do everything the Aspell utility can do and not just be able spell
4144 * Develop a *more powerful C API* for Aspell. Ideally this API
4145 should allow one to perform all the tasks the Aspell utility can
4146 do. This included the ability to check whole documents, and create
4147 dictionaries, among other things.
4149 * Create a *C++ interface* for Aspell, possibly on top of the C one.
4153 File: aspell.info, Node: Other Items, Next: Notes on Various Items, Prev: Important Items, Up: To Do
4158 These items all sound like good ideas however I am not sure when I will
4159 get to implementing them if ever. Words in bold indicate how you
4160 should refer to the item when discussing it with me or others.
4162 * Come up with a plug-in for `gEdit' the gnome text editor.
4164 * Change languages (and thus dictionaries) based on the information
4165 in the actual document.
4167 * Come up with a mode that will skip words based on the symbols that
4168 (almost) always surround the word. *Note Word skipping by
4171 * Create two *server modes* for Aspell. One that uses the DICT
4172 protocol and one that uses `ispell -a' method of communication via
4173 some arbitrary port.
4175 * Come up with *thread safe personal dictionaries*.
4177 * Use the *Hidden Markov Model* to base the suggestions on not only
4178 the word itself but on the context around the word. *Note Hidden
4181 * Having a way to *email the personal dictionary* and/or replacement
4182 list to a particular address either periodically or when it grows
4183 to a certain size. *Note Email the Personal Dictionary::.
4185 The following good ideas were found in the Ispell `WISHES' file so I
4186 thought I would pass them on.
4188 * Ispell should be smart enough to ignore hyphenation signs, such as
4189 the TeX `\-' hyphenation indicator.
4191 * (Jeff Edmonds) The personal dictionary should be able to remove
4192 certain words from the master dictionary, so that obscure words
4193 like "wether" wouldn't mask favorite typos.
4195 * (Jeff Edmonds) It would be wonderful if Ispell could correct
4196 inserted spaces such as "th e" for "the" or even "can not" for
4199 * Since Ispell has dictionaries available to it, it is conceivable
4200 that it could automatically determine the language of a particular
4201 file by choosing the dictionary that produced the fewest spelling
4202 errors on the first few lines.
4205 File: aspell.info, Node: Notes on Various Items, Prev: Other Items, Up: To Do
4207 D.3 Notes on Various Items
4208 ==========================
4212 * Word skipping by context::
4213 * Hidden Markov Model::
4214 * Email the Personal Dictionary::
4217 File: aspell.info, Node: Word skipping by context, Next: Hidden Markov Model, Up: Notes on Various Items
4219 D.3.1 Word skipping by context
4220 ------------------------------
4222 This was posted on the Aspell mailing list on January 1, 1999:
4224 I had an idea on a great general way to determine if a word should be
4225 skipped. Determine the words to skip based on the symbols that
4226 (almost) always surround the word.
4228 For example when asked to check the following C++ code:
4229 cout << "My age is: " << num << endl;
4230 cout << "Next year I will be " << num + 1 << endl;
4232 `cout', `num', and `endl' will all be skipped. `cout' will be
4233 skipped because it is always preceded by a `<<'. `num' will be skipped
4234 because it is always preceded by a `<<'. And `endl' will be skipped
4235 because it is always between a `<<' and a `;'.
4237 Given the following HTML code.
4238 <table width=50% cellspacing=0 cellpadding=1>
4239 <tr><td>One<td>Two<td>Three
4243 <table cellspacing=0 cellpadding=1>
4246 `table', `width' `cellspacing', `cellpadding', `tr', `td' will all
4247 be skipped because they are always enclosed in `<>'. Now of course
4248 `table' and `width' would be marked as correct anyway however there is
4249 no harm in skipping them.
4251 So I was wondering if anyone on this list has any experience in
4252 writing this sort of context recognition code or could give me some
4253 pointers in the right direction.
4255 This sort of word skipping will be very powerful if done right. I
4256 imagine that it could replace specific spell checker modes for TeX,
4257 Nroff, SGML etc because it will automatically be able to figure out
4258 where it should skip words. It could also probably do a very good job
4259 on programming languages code.
4261 If you are interested in helping me out with this or just have
4262 general comments about the idea please let me know.
4265 File: aspell.info, Node: Hidden Markov Model, Next: Email the Personal Dictionary, Prev: Word skipping by context, Up: Notes on Various Items
4267 D.3.2 Hidden Markov Model
4268 -------------------------
4270 Knud Haugaard Sørensen suggested this one. From his email on the
4271 Aspell mailing list:
4273 consider these examples:
4281 the example illustrates that the right correction might depend on the
4282 context of the word.
4284 So I suggested that you take a look on HMM to solve this problem.
4286 This might also provide a good base to include grammar correction in
4289 see this link `http://www.cse.ogi.edu/CSLU/HLTsurvey/ch1node7.html'.
4291 I think it is a great idea. However unfortunately it will probably
4292 be very complicated to implement. Perhaps in the far future.
4295 File: aspell.info, Node: Email the Personal Dictionary, Prev: Hidden Markov Model, Up: Notes on Various Items
4297 D.3.3 Email the Personal Dictionary
4298 -----------------------------------
4300 Someone suggested in a personal email:
4302 Have you thought of adding a function to Aspell, that - when the
4303 personal dictionary has grown significantly - sends the user's
4304 personal dictionary to the maintainer of the corresponding Aspell
4305 dictionary? (if the user allows it)
4307 It would be a very useful service to the dictionary maintainers,
4308 and I think most users can see their benefit in it too.
4312 Yes I have considered something like that but not for the personal
4313 dictionaries but rather the replacement word list in order to get
4314 better test data for `http://aspell.sourceforge.net/test/'.
4316 The problem is I don't know of a good way to do this since Aspell can
4317 also be used as a library. It also is not a real high priority,
4318 especially since I would first need to learn how to send email within a
4322 File: aspell.info, Node: Installing, Next: ChangeLog, Prev: To Do, Up: Top
4324 Appendix E Installing
4325 *********************
4327 Aspell requires gcc 2.95 (or better) as the C++ compiler. Other C++
4328 compilers should work with some effort. Other C++ compilers for mostly
4329 POSIX compliant (Unix, Linux, BeOS, Cygwin) systems should work without
4330 any major problems provided that the compile can handle all of the
4331 advanced C++ features Aspell uses. C++ compilers for non-Unix systems
4332 might work but it will take some work. Aspell at very least requires a
4333 Unix-like environment (`sh', `grep', `sed', `tr', ...), and Perl in
4334 order to build. Aspell also uses a few POSIX functions when necessary.
4336 The latest version can always be found at GNU Aspell's home page at
4337 `http://aspell.net'.
4341 * Generic Install Instructions::
4342 * HTML Manuals and "make clean"::
4344 * Loadable Filter Notes::
4345 * Upgrading from Aspell 0.50::
4346 * Upgrading from Aspell .33/Pspell .12::
4347 * Upgrading from a Pre-0.50 snapshot::
4351 File: aspell.info, Node: Generic Install Instructions, Next: HTML Manuals and "make clean", Up: Installing
4353 E.1 Generic Install Instructions
4354 ================================
4358 For additional `configure' options type `./configure --help'. You
4359 can control what C++ compiler is used by setting the environment
4360 variable `CXX' before running configure and you can control what flags
4361 are passed to the C++ compile via the environment variable `CXXFLAGS'.
4362 Static libraries are disabled by default since static libraries will
4363 not work right due to the mixing of C and C++. When a C program links
4364 with the static libraries in Aspell it is likely to crash because
4365 Aspell's C++ objects are not getting initialized correctly. However,
4366 if for some reason you want them, you can enable them via
4369 Aspell should then compile without any additional user intervention.
4370 If you run into problems please first check the sections below as that
4371 might solve your problem.
4373 To install the program simply type
4377 After Aspell is installed at least one dictionary needs to be
4378 installed. You can find them at `http://aspell.net/'. The `aspell'
4379 program must be in your path in order for the dictionaries to install
4382 If you do not have Ispell or the traditional Unix `spell' utility
4383 installed on your system then you should also copy the compatibility
4384 scripts `ispell' and `spell' located in the `scripts/' directory into
4385 your binary directory which is usually `/usr/local/bin' so that
4386 programs that expect the `ispell' or `spell' command will work
4390 File: aspell.info, Node: HTML Manuals and "make clean", Next: Curses Notes, Prev: Generic Install Instructions, Up: Installing
4392 E.2 HTML Manuals and `make clean'
4393 =================================
4395 The Aspell distribution includes HTML versions of the User and
4396 Developer's manual. Unfortunately, doing a `make clean' will erase
4397 them. This is due to a limitation of automake which is not easily
4398 fixed. If makeinfo is installed they can easily be rebuild with `make
4399 aspell.html aspell-dev.html', or you can unpack them from the tarbar.
4402 File: aspell.info, Node: Curses Notes, Next: Loadable Filter Notes, Prev: HTML Manuals and "make clean", Up: Installing
4407 If you are having problems compiling `check_funs.cpp' then the most
4408 likely reason is due to incompatibilities with the curses
4409 implementation on your system. You should first try disabling the
4410 "wide" curses library by with the `--disable-wide-curses' configure
4411 option.. By doing so you will lose support for properly displaying
4412 UTF-8 characters but you may still be able to get the full screen
4413 interface. If this fails than you can disable curses support
4414 altogether with the `--disable-curses' configure option. By doing this
4415 you will lose the nice full screen interface but hopefully you will be
4416 able to at least get Aspell to compile correctly.
4418 If the curses library is installed in a non-standard location than
4419 you can specify the library and include directory with
4420 `--enable-curses=LIB' and `--enable-curses-include=DIR'.
4422 `LIB' can either be the complete path of the library--for example
4423 /usr/local/curses/libcurses.a
4424 or the name of the library (for example `ncurses') or a combined
4425 location and library in the form `-LLIBDIR -lLIB' (for example
4426 `-L/usr/local/ncurses/lib -lncurses').
4428 DIR is the location of the curses header files (for example
4429 `/usr/local/ncurses/include').
4431 E.3.1 Unicode Support
4432 ---------------------
4434 In order for Aspell to correctly spell check UTF-8 documents in full
4435 screen mode the "wide" version of the curses library must be installed.
4436 This is different from the normal version of curses library, and is
4437 normally named `libcursesw' (with a `w' at the end) or `libncursesw'.
4438 UTF-8 documents will not display correctly without the right curses
4441 In addition your system must also support the `mblen' function.
4442 Although this function was defined in the ISO C89 standard (ANSI
4443 X3.159-1989), not all systems have it.
4446 File: aspell.info, Node: Loadable Filter Notes, Next: Upgrading from Aspell 0.50, Prev: Curses Notes, Up: Installing
4448 E.4 Loadable Filter Notes
4449 =========================
4451 Support for being able to load additional filter modules at run-time
4452 has only been verified to work on Linux platforms. If you get linker
4453 errors when trying to use a filter, then it is likely that loadable
4454 filter support is not working yet on your platform. Thus, in order to
4455 get Aspell to work correctly you will need to avoid compiling the
4456 filters as individual modules by using the
4457 `--enable-compile-in-filters' when configuring Aspell with
4461 File: aspell.info, Node: Upgrading from Aspell 0.50, Next: Upgrading from Aspell .33/Pspell .12, Prev: Loadable Filter Notes, Up: Installing
4463 E.5 Upgrading from Aspell 0.50
4464 ==============================
4466 The dictionary format has changed so dictionaries will need to be
4469 All data, by default, is now included in `LIBDIR/aspell-0.60' so
4470 that multiple versions of Aspell can more peacefully coexist. This
4471 included both the dictionaries and the language data files which were
4472 stored in `SHAREDIR/aspell' before Aspell 0.60.
4474 The format of the character data files has changed. The new
4475 character data files are installed with Aspell so you should not have
4476 to worry about it unless you made a custom one.
4478 The dictionary option `strip-accents' has been removed. For this
4479 reason the old English dictionary (up to 0.51) will no longer work. A
4480 new English dictionary is now available which avoids using this option.
4481 In addition the `ignore-accents' option is currently unimplemented.
4483 The flag `-l' is now a shortcut for `--lang', instead of `--list' as
4484 it was with Aspell 0.50.
4486 E.5.1 Binary Compatibility
4487 --------------------------
4489 The Aspell 0.60 library is binary compatible with the Aspell 0.50
4490 library. For this reason I chose _not_ to increment the major version
4491 number (so-name) of the shared library by default which means programs
4492 that were compiled for Aspell 0.50 will also work for Aspell 0.60.
4493 However, this means that having both Aspell 0.50 and Aspell 0.60
4494 installed at the same time can be pragmatic. If you wish to allow both
4495 Aspell 0.50 and 0.60 to be installed at the same time then you can use
4496 the configure option `--incremented-soname' which will increment
4497 so-name. You should only use this option if you know what you are
4498 doing. It is up to you to somehow ensure that both the Aspell 0.50 and
4499 0.60 executables can coexist.
4501 If after incrementing the so-name you wish to allow programs compiled
4502 for Aspell 0.50 to use Aspell 0.60 instead (thus implying that Aspell
4503 0.50 is not installed) then you can use a special compatibility library
4504 which can be found in the `lib5' directory. This directory will not be
4505 entered when building or installing Aspell so you must manually build
4506 and install this library. You should build it after the rest of Aspell
4507 is built. The order in which this library is installed, with relation
4508 to the rest of Aspell, is also important. If it is installed _after_
4509 the rest of Aspell then new programs will link to the old library
4510 (which will work for Aspell 0.50 or 0.60) when built, if installed
4511 _before_, new programs will link with the new library (Aspell 0.60
4515 File: aspell.info, Node: Upgrading from Aspell .33/Pspell .12, Next: Upgrading from a Pre-0.50 snapshot, Prev: Upgrading from Aspell 0.50, Up: Installing
4517 E.6 Upgrading from Aspell .33/Pspell .12
4518 ========================================
4520 Aspell has undergone an extremely large number of changes since the
4521 previous Aspell/Pspell release. For one thing Pspell has been merged
4522 with Aspell so there in no longer two separate libraries you have to
4525 Because of the massive changes between Aspell/Pspell and Aspell 0.50
4526 you may want to clean out the old files before installing the the new
4527 Aspell. To do so do a `make uninstall' in the original Aspell and
4528 Pspell source directories.
4530 The way dictionaries are handled has also changed. This includes a
4531 change in the naming conventions of both language names and
4532 dictionaries. Due to the language name change, your old personal
4533 dictionaries will not be recognized. However, you can import the old
4534 dictionaries by running the `aspell-import' script. This also means
4535 that dictionaries designed to work with older versions of Aspell are
4536 not likely to function correctly. Fortunately new dictionary packages
4537 are available for most languages. You can find them off of the Aspell
4538 home page at `http://aspell.net'.
4540 The Pspell ABI is now part of Aspell except that the name of
4541 everything has changed due to the renaming of Pspell to Aspell. In
4542 particular please note the following name changes:
4546 emulation -> enumeration
4547 master_word_list -> main_word_list
4549 Please also note that the name of the `language-tag' option has
4550 changed to `lang'. However, for backward compatibility the
4551 `language-tag' option will still work.
4553 However, you should also be able to build applications that require
4554 Pspell with the new Aspell as a backward compatibility header file is
4557 Due to a change in the way dictionaries are handled, scanning for
4558 `.pwli' files in order to find out which dictionaries are available
4559 will no longer work. This means that programs that relied on this
4560 technique may have problems finding dictionaries. Fortunately, GNU
4561 Aspell now provided a uniform way to list all installed dictionaries
4562 via the c API. See the file `list-dicts.c' in the `examples/'
4563 directory for an example of how to do this. Unfortunately there isn't
4564 any simple way to find out which dictionaries are installed which will
4565 work with both the old Aspell/Pspell and the new GNU Aspell.
4568 File: aspell.info, Node: Upgrading from a Pre-0.50 snapshot, Next: WIN32 Notes, Prev: Upgrading from Aspell .33/Pspell .12, Up: Installing
4570 E.7 Upgrading from a Pre-0.50 snapshot
4571 ======================================
4573 At the last minute I decided to merge the `speller-util' program into
4574 the main `aspell' program. You may wish to remove that `speller-util'
4575 program to avoid confusion. This also means that dictionaries designed
4576 to work with the snapshot will no longer work with the official release.
4579 File: aspell.info, Node: WIN32 Notes, Prev: Upgrading from a Pre-0.50 snapshot, Up: Installing
4584 E.8.1 Getting the WIN32 version
4585 -------------------------------
4587 The latest version of the native Aspell/WIN32 port, including binaries,
4588 can be found at `http://aspell.net/win32'. This page has,
4589 unfortunately, not been updated for Aspell 0.60. If you are interested
4590 in updated the native port please let me know.
4592 E.8.2 Building the WIN32 version
4593 --------------------------------
4595 There are two basically different ways of building Aspell using GCC for
4596 WIN32: You can either use the Cygwin compiler, which will produce
4597 binaries that depend on the POSIX layer in `cygwin1.dll'. The other
4598 way is using MinGW GCC, those binaries use the native C runtime from
4599 Microsoft (MSVCRT.DLL).
4601 E.8.2.1 Building Aspell using Cygwin
4602 ....................................
4604 This works exactly like on other POSIX compatible systems using the
4605 `./configure && make && make install' cycle. Some versions of Cygwin
4606 GCC will fail to link, this is caused by an incorrect `libstdc++.la' in
4607 the `/lib' directory. After removing or renaming this file, the build
4608 progress should work (GCC-2.95 and GCC-3.x should work).
4610 E.8.2.2 Building Aspell using MinGW
4611 ...................................
4613 There are several different ways to build Aspell using MinGW. The
4614 easiest way is to use a Cygwin compiler but instruct it to build a
4615 native binary rather than a Cygwin one. To do this configure with:
4617 ./configure CFLAGS='-O2 -mno-cygwin' CXXFLAGS='-O2 -mno-cygwin'
4619 You may also want to add the option `--enable-win32-relocatable' to
4620 use more windows friendly directories. *Note Win32-Directories::. In
4621 this case configure with:
4623 ./configure CFLAGS='-O2 -mno-cygwin' CXXFLAGS='-O2 -mno-cygwin' --enable-win32-relocatable
4625 It should also be possible to build Aspell using the MSYS
4626 environment. But this has not been very well tested. If building with
4627 MSYS _do not_ add `CFLAGS ...' to configure.
4629 E.8.2.3 Building Aspell without using Cygwin or MSYS
4630 ....................................................
4632 It is also possible to build Aspell without Cygwin of MinGW by using
4633 the files in the `win32/' subdirectory. However, these files have not
4634 been updated to work with Aspell 0.60. Thus the following instructions
4635 will not work without some effort. If you do get Aspell to compile
4636 this way please send me the updated files so that I can include them
4637 with the next release.
4639 To compile Aspell with the MinGW compiler, you will need at least
4640 GCC-3.2 (as shipped with MinGW-2.0.3) and some GNU tools like `rm' and
4641 `cp'. The origin of those tools doesn't matter, it has shown to work
4642 with any tools from MinGW/MSys, Cygwin or Linux. To build Aspell, move
4643 into the `win32' subdirectory and type `make'. You can enable some
4644 additional build options by either commenting out the definitions at
4645 the head of the Makefile or passing those values as environment
4646 variables or at the `make' command line. Following options are
4650 If set to "1", the binaries will include debugging information
4651 (resulting in a much bigger size).
4654 Enter the path to the pdcurses library here, in order to get a
4655 nicer console interface (see below).
4658 Enter the filename of MS `lib.exe' here, if you want to build
4659 libraries that can be imported from MS Visual C++.
4662 If set to "1", Aspell will detect the prefix from the path where
4663 the DLL resides (see below for further details).
4666 Sets a prefix to be used for cross compilation (e.g.
4667 `/usr/local/bin/i586-mingw32msvc-' to cross compile from Linux).
4669 There are also a MinGW compilers available for Cygwin and Linux, both
4670 versions are able to compile Aspell using the prebuilt `Makefile'.
4671 While the Cygwin port automatically detects the correct compiler, the
4672 Linux version depends on setting the `TARGET' variable in the
4673 `Makefile' (or environment) to the correct compiler prefix.
4675 Other compilers may work. There is a patch for MS Visual C++ 6.0
4676 available at `ftp://ftp.gnu.org/gnu/aspell', but it needs a lot of
4677 changes to the Aspell sources. It has also been reported that the
4678 Intel C++ compiler can be used for compilation.
4683 In order to get the nice full screen interface when spell checking
4684 files, a curses implementation that does not require Cygwin is
4685 required. The PDCurses (`http://pdcurses.sourceforge.net')
4686 implementation is known to work, other implementations may work however
4687 they have not been tested. See the previous section for information on
4688 specifying the location of the curses library and include file.
4692 * PDcurses built with MinGW needs to be compiled with
4693 `-DPDC_STATIC_BUILD' to avoid duplicate declaration of `DllMain'
4694 when compiling `aspell.exe'.
4696 * The curses enabled version can cause trouble in some shells (MSys
4697 `rxvt', `emacs') and will produce errors like `initscr() LINES=1
4698 COLS=1: too small'. Use a non-curses version for those purposes.
4703 If Aspell is configured with `--enable-win32-relocatable' or compiled
4704 with `WIN32_RELOCATABLE=1' when using a Makefile, it can be run from
4705 any directory: it will set `PREFIX' according to its install location
4706 (assuming it resides in `PREFIX\\bin'). Your personal wordlists will
4707 be saved in the `PREFIX' directory with their names changed from
4708 `.aspell.LANG.*' to `LANG.*' (you can override the path by setting the
4709 `HOME' environment variable).
4714 The installer registers the DLLs as shared libraries, you should
4715 increase the reference counter to avoid the libraries being uninstalled
4716 if your application still depends on them (and decrease it again when
4717 uninstalling your program). The reference counters are located under:
4718 HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion\SharedDLLs
4720 The install location and version numbers are stored under
4722 HKLM\SOFTWARE\Aspell
4724 E.8.6 WIN32 consoles
4725 --------------------
4727 The console uses a different encoding than GUI applications, changing
4728 this to to a Windows encoding (e.g. 1252) is not supported on
4729 Win9x/Me. On WinNT (and later) those codepages can be set by first
4730 changing the console font to `lucida console', then changing the
4731 codepage using `chcp 1252'.
4733 Some alternative shells (e.g. MSys' `rxvt' or Cygwin's `bash') do a
4734 codepage conversion (if correctly set up), so running Aspell inside
4735 those shells might be a workaround for Win9x.
4738 File: aspell.info, Node: ChangeLog, Next: Authors, Prev: Installing, Up: Top
4740 Appendix F ChangeLog
4741 ********************
4743 Changes from 0.60.6 to 0.60.6.1 (July 4, 2011)
4744 ==============================================
4746 * Update to Automake 1.10.3.
4748 * Fix a bug which caused a race condition (leading to a likely crash)
4749 when two threads try to update the dictionary cache at the same
4752 * Make it very clear that compiling Aspell with NDEBUG is a bad idea
4753 (see `http://aspell.net/ndebug.html') by outputting a warning when
4754 building with NDEBUG defined.
4756 * Numerous other minor updates and bug fixes.
4758 Changes from 0.60.5 to 0.60.6 (April 16, 2007)
4759 ==============================================
4761 * Compile fixes for Gcc 4.3.
4763 * Updated to Libtool 2.2.2 and Automake 1.10.1
4765 * Minor tweak to suggestion code which improved suggestion results in
4768 * Always line buffer stdout and stderr in the Aspell utility when
4769 there is the potential for it to be used interactively through a
4772 * Removed debug output in `aspell munch-list'.
4774 * Other minor updates and bug fixes.
4776 Changes from 0.60.4 to 0.60.5 (December 18, 2006)
4777 =================================================
4779 * Compile fix for Gcc 4.1
4781 * Updated to Gettext 0.16.1, Libtool 1.5.22, Automake 1.10, Autoconf
4784 * Documentation improvements, including an updated `man' page.
4786 * Complain if more than one file is specified when checking files
4787 using the `aspell check' command, rather than ignoring the other
4790 * Large number of bug fixes.
4792 Changes from 0.60.3 to 0.60.4 (October 19, 2005)
4793 ================================================
4795 * Fixed a bug that caused Aspell to crash when checking certain
4796 Russian words, this bug likely affected other languages as well.
4798 * Updated to Gettext 0.14.5 which is required for AMD64, also
4799 updated to to Libtool 1.5.20.
4801 * Fixed an alignment bug which caused mmap to always fail when
4802 reading in dictionaries.
4804 * Added note about how `make clean' will remove the HTML manuals.
4806 * Added manual page for prezip-bin and enhanced word-list-compress
4807 manual page thanks to the work of Jose Da Silva.
4809 * Other minor updates and bug fixes.
4811 Changes from 0.60.2 to 0.60.3 (June 28, 2005)
4812 =============================================
4814 * Fixed bugs involving several of the C API functions.
4816 * Fixed bug where `ultra' or `fast' mode would not return any
4817 suggestions when soundslike lookup was not used.
4819 * Made a minor, yet significant, optimization to the suggestion code.
4820 This sped things up by an order of magnitude in some cases.
4822 * Avoid using the slow ngram scan except when the `sug-mode' is
4823 `slow' or `bad-speller'.
4825 * Fixed a bug in curses mode which caused word-wrap to not work
4826 correctly in some cases.
4828 * Fixed a bug in pipe mode with a missing newline.
4830 * Fixed the `spell' compatibility script.
4832 * Several other minor bugs fixed.
4834 * Made note about the change in behavior of the `-l' command line
4837 * Other manual update/fixes.
4839 * Updated to Libtool 1.5.18, Automake 1.9.6, and Makeinfo 4.8.
4841 Changes from 0.60.1 to 0.60.2 (December 18, 2004)
4842 =================================================
4844 * Added the `munch-list' command to the Aspell utility. The `munch'
4845 program in the `myspell/' directory will disappear in Aspell 0.61.
4846 The `munchlist' script will also likely disappear or be replaced
4847 when Aspell 0.61 is released since it doesn't work correctly
4850 * Several important bug fixes some of which rendered some non-English
4853 * Other minor changes.
4855 Changes from 0.60.1 to 0.60.1.1 (November 20, 2004)
4856 ===================================================
4858 * Fix bug involving checking of capitalized word when affix
4859 compression is used.
4863 * Added an option to disable using the "wide" curses version in case
4864 it causes compile problems.
4866 * Minor manual updates
4868 * Avoided including some unnecessary files in the distribution.
4870 Changes from 0.60 to 0.60.1 (November 7, 2004)
4871 ==============================================
4873 * Lots of compile fixes for various platforms.
4875 * Miscellaneous bug fixes.
4877 * Added Nroff filter thanks to Sergey Poznyakoff.
4879 * The default filter mode when in pipe mode is now nroff for
4880 compatibility with Ispell.
4882 * Added Texinfo filter.
4884 * Added a section detailing the differences between Ispell and
4887 * Updated the section on thread safety.
4889 * Other miscellaneous manual changes such as updating the To Do and
4892 Changes from 0.50.5 to 0.60 (August 27, 2004)
4893 =============================================
4895 * Added support for Affix Compression. Affix compression stores the
4896 root word and then a list of prefixes and suffixes that the word
4897 can take, and thus saves a lot of space. The codebase comes from
4898 MySpell found in OpenOffice. It uses the same affix file that
4899 OpenOffice (and Mozilla) use. Affix compression will even work
4900 with soundslike lookup to a limited extent.
4902 * Added support for accepting all input and printing all output in
4903 UTF-8 or some other encoding different from the one Aspell uses.
4904 This includes support for Unicode normalization. Aspell can now
4905 support any language with no more than 210 distinct characters,
4906 including different capitalizations and accents, _even if_ there
4907 is not an existing 8-bit encoding that supports the language.
4909 * Added support for loadable filters and customizable filter modes
4910 thanks to Christoph Hintermüller.
4912 * Enhanced SGML filter to also support skipping sgml tags such as
4913 "script" blocks thanks to Tom Snyder.
4915 * Added gettext support thanks to Sergey Poznyakoff
4917 * Reworked the compiled dictionary format. Compiled dictionaries
4918 now take up less space (less than 80% for the English language) and
4919 creating them is significantly faster (over 4 times for the
4922 * Reworked suggestion code. It is significantly faster when dealing
4923 with short words (up to 10 times). Also added support for MySpell
4924 Replacement Tables and n-gram lookup. In addition, added basic
4925 support for compound words.
4927 * Manual has has been converted to texinfo format thanks to the work
4930 * Reworked the build system so that a single Makefile is used for
4933 * All data, by default, is now included in `LIBDIR/aspell-0.60'.
4934 Also added a built time option to increment the major version
4935 number of the shared library. This should allow both Aspell
4936 version 0.50 and 0.60 to coexist. The major version number is
4937 _not_ incremented by default as Aspell 0.60 is binary compatible
4938 with Aspell 0.50. *Note Binary Compatibility::.
4940 * The code to handle dictionaries has been rewritten. Because of
4941 this support for the dictionary option `strip-accents' has been
4942 removed. In addition the `ignore-accents' option is currently
4945 * Lots of other minor changes due to massive overhaul of the source
4949 Changes from 0.50.4.1 to 0.50.5 (Feb 10, 2004)
4950 ==============================================
4952 * Reworked url filter which fixed several bugs and now accepts
4953 "bla.bla/kdkdl" as a url.
4955 * Fixed bug in which the url filter was coming before all other
4956 filters when it was supposed to come after. This solved a number
4957 of problems where the url filter was interfering with other
4960 * Small bug fix in SGML filter.
4962 * Added code page charsets, ie cp125?.dat.
4964 * Added natural (split) keyboard data file as "split.kbd"
4966 * Compile fixes for the upcoming Gcc 3.4
4968 * Removed Solaris link hack as it was causing more problems than it
4971 * Compile fixes for Sun WorkShop 6 compiler, but there may still be
4972 some problems, especially with linking.
4974 * Included patch to help compile with Microsoft Visual C++ 6.
4976 * Minor manual fixes.
4978 * Updated the TODO section to reflect the current progress with the
4979 next major version of Aspell (0.51).
4981 * Updated to Autoconf 2.59, Automake 1.82, and Libtool 1.5.2.
4983 Changes from 0.50.4 to 0.50.4.1 (Oct 11, 2003)
4984 ==============================================
4986 * Fixed major bug in pipe mode which caused the last character to be
4987 chopped off words before they were stored.
4989 * Minor formating fixes in the manual.
4991 Changes from 0.50.3 to 0.50.4 (Sep 26, 2003)
4992 ============================================
4994 * Minor changes in URL filter to avoid treating the double quote
4995 character as part of the URL, and to avoid treating words ending in
4996 more than one period as a URL.
4998 * Document fixes in Aspell API
5000 * Small compile fixes, including one for GCC 3.3
5002 * Updated Win32 section since a port now exists thanks to Thorsten
5005 * Complain instead of doing nothing or aborting for unimplemented
5006 functions in Aspell utility.
5008 * Portability bug fixes.
5010 * Upgraded to Autoconf 2.57, Automake 1.7.7, Libtool 1.5 (no longer
5011 use CVS version of libtool).
5013 Changes from 0.50.2 to 0.50.3 (Nov 23, 2002)
5014 ============================================
5016 * Hopefully fixed the Ispell alignment error problem when Aspell is
5017 used with ispell.el.
5019 * Fixed a problem with personal dictionaries on NFS mounted home
5022 * Compiled libaspell-common directory into libaspell for now to avoid
5023 forcing applications to relink whenever a new Aspell version is out
5024 which was due to the use of the libtool '-release' flag.
5026 * Fixed Makefiles so that Aspell can be built outside the source tree
5029 * Updated the section on compiling with Win32.
5031 * Updated to Autoconf 2.56.
5033 Changes from 0.50.1 to 0.50.2 (Sep 28, 2002)
5034 ============================================
5036 * Fixed a number of bugs in Ispell compatibility mode
5038 * Fixed a number of bugs with the handling of replacement pairs
5040 * Other miscellaneous bug fixes
5042 * Additional Win32 portability fixes
5044 * Added the Ukrainian KOI8-U charset.
5046 Changes from 0.50 to 0.50.1 (Aug 28, 2002)
5047 ==========================================
5049 * A rather large number of portability fixes for non GNU/Linux
5052 * Fixed pkglibdir and pkgdatadir in configure.
5054 * Reintroduced some configure options from Aspell .33.7 included
5055 dict-dir, data-dir, curses, curses-include, win32-relocatable.
5057 * Fixed Aspell so it will now compile with -O3 when using gcc.
5059 * Updated note on Win32 support.
5061 * Other minor manual improvements.
5063 * Portability fixes in dictionary files
5065 * Official dictionary package for the Slovak language.
5067 Changes from .33.7.1 to 0.50 (Aug 23, 2002)
5068 ===========================================
5070 * A complete overhaul of the source code which included merging
5073 * Changed the way dictionaries and languages are handled.
5075 * Added Dvorak keymap.
5077 * Added the ability to list the available dictionaries
5079 * Improved the spell checking interface a bit.
5081 * Added support for using the Ispell keymapping when checking files.
5083 * Complete rewrite of the filter interface. It should now be
5084 fairly easy to add new filters to Aspell.
5086 * Added some preliminary developer documentation.
5088 * Lots of other changes due to the massive overhaul of the source
5091 Changes from .33.7 to .33.7.1 (Aug 20, 2001)
5092 ============================================
5094 * Minor manual fixes.
5096 * Compile fix for Gcc 3.0 and Solaris.
5098 Changes from .33.6.3 to .33.7 (Aug 2, 2001)
5099 ===========================================
5101 * Updates to Autoconf 2.50 and switched to the HEAD branch of
5104 * Fixed a bug which caused Aspell to crash when typo-analysis was not
5105 used such as when sug-mode is *fast* or *bad spellers*.
5107 * Added support for typo-analysis even when a soundslike was not
5110 * Fixed a bug which causes extended charters to display incorrectly
5113 * Compile fixes so that it will compile with Gcc 3.0.
5115 * Compile fixed which should allow Aspell to compile with Egcs 1.1.
5116 I have not been able to actually test it though. Please let me
5117 know at kevina@users.sourceforge.net if you have tried with Egcs
5120 * Compile and configuration script fixes so that USE_FILE_INO will
5121 properly be defined and Aspell will compile correctly when it is
5124 * More ANSI C++ compliance fixes.
5126 Changes from .33.6.2 to .33.6.3 (June 3, 2001)
5127 ==============================================
5129 * Fixed a build problem in the manual/ directory by including
5130 manual-text and manual-html in the distribution.
5132 Changes from .33.6.1 to .33.6.2 (June 3, 2001)
5133 ==============================================
5135 * Compile fix so that Aspell will work correctly when not installed
5138 * Avoided regenerating the manual unless configured with
5139 enable-maintainer-mode.
5141 * Added the missing documentation files in the scowl directory.
5143 Changes from .33.6 to .33.6.1 (May 29, 2001)
5144 ============================================
5146 * Fixed a formating problem with the manual involving <.
5148 * Added a note about creating pwli files.
5150 * Removed the space after between the -L and the directory name in
5151 the pspell-module/Makefile which caused problems on some
5154 * Added the configure option AM_MAINTAINER_MODE to avoid enabling
5155 rules which often causes generated build files to be rebuilt with
5156 the wrong version of Libtool by default. I don't know why I
5157 didn't think to do this a long time ago.
5159 Changes from .33.5 to .33.6 (May 18, 2001)
5160 ==========================================
5162 * Fixed a minor bug where some words would have random compound tags
5165 * Fixed a compile problem on many platforms where fileno is defined
5168 * Updated the description for a few of Aspell's options.
5170 * Removed the note of Aspell not being able to run when compiled with
5171 the upcoming Gcc 3.0 compiler as things seam to work now.
5173 * Added a note about Aspell not being able to compile with Egcs 1.1.
5175 * Added hack to deal with Libtool's interdependencies problem. See
5176 bug #416981 for Pspell for more info.
5178 Changes from .33 to .33.5 (April 5, 2001)
5179 =========================================
5181 * *dump master* correctly detects which dictionary and language to
5182 use based on the `LANG' environment variable.
5184 * Fixed a problem on Win32 which involves path names that began with
5187 * Bug fixes and enhancements so that Aspell can once again run under
5188 MinGW. You can even use the new full screen interface if Aspell is
5189 compiled with PDCurses.
5191 * Some major modifications to make Aspell more C++ compliant in
5192 order to get Aspell to compile under the upcoming Gcc 3.0
5193 compiler. This included only using STL features found in the
5194 standard version of C++. (Which means Aspell will no longer
5195 require using the SGI version of the STL) This should also make
5196 compiling C++ under non-gcc compilers a lot simpler. Please note
5197 that Aspell still has some problems with the upcoming Gcc 3.0
5200 * Minor changes to remove some -Wall warnings.
5202 * Added a hack so that Aspell would properly compile as a shared
5203 library under Solaris.
5205 * Added a few important missing words to the English word list.
5207 Changes from .32.6 to .33 (January 28, 2001)
5208 ============================================
5210 * Added a new new curses based interface to replace the dumb terminal
5211 interface everyone has been bitching about.
5213 * Added the ability to give higher priority to words such as "the"
5214 instead of "teh" which are likely to be due to typos.
5216 * Reorganized the manual so that it is hopefully easier to follow.
5218 * Ability to automatically select the best dictionary to use based on
5219 the setting of the `LANG' environment variable.
5221 * Expanded the medium dictionary size to include more words which
5222 included the original words found in Ispell and eliminated the
5225 * Added three special variant add-on dictionaries.
5227 * Switched to the multi-language branch of the CVS version of
5231 Changes from .32.5 to .32.6 (Nov 8, 2000)
5232 =========================================
5234 * Fixed a bug where Aspell would crash when reading-in accented
5235 characters on some platforms. This fixed bug # 112435.
5237 * Fixed some other bugs so that it will run under Win32 under CygWin.
5238 Unfortunately it still won't run properly under Mingw.
5240 * Fixed the mmap test in configure so that it won't fail on some
5241 platforms that use munmap(char *, int) instead of munmap(void *,
5244 * Upgraded to the latest CVS version of libtool which fixed the
5245 problem with using GNU Make under Solaris.
5247 * Added an option to copy files instead of using symbolic links for
5248 the special *multi* dictionary files.
5250 Changes from .32.1 to .32.5 (August 18, 2000)
5251 =============================================
5253 * Changed my email from kevinatk at home com to kevina at users
5254 sourceforge net please make a note of the new email address.
5256 * Added an option to control if the personal replacement dictionary
5257 is saved when the save_all_wls method is called.
5259 * Brought back the ability to dump the master word list even in the
5260 case of the special *multi* lists.
5262 * Added a large number of hacker related words and some other slang
5263 terms to the medium size word list.
5265 * Added an *ispell* and *spell* compatibility script for systems
5266 which don't have Ispell installed. They are located in the
5267 scripts/ directory and are not installed by default.
5271 * Added a note on not using GNU Make on Solaris.
5273 Changes from .32 to .32.1 (August 5, 2000)
5274 ==========================================
5276 * Minor compile fixes for recent gcc snapshot.
5278 * Fixed naming of pwli files.
5280 * Fixed a bug when Aspell will crash when used with certain single
5281 letter flags. This bug was most noticeable when used with Emacs.
5283 * Word list changes, see SCOWL Readme.
5285 * Other miscellaneous changes.
5287 Changes from .31.1 to .32 (July 23, 2000)
5288 =========================================
5290 * Added support for optionally doing without the soundslike data.
5292 * Greatly reduced the amount of memory used when creating word lists.
5294 * Added support for ignoring accents when coming up with suggestions.
5296 * Added support for local-data-dir which is searched before data-dir.
5298 * Added support for specifying which words may be used in compounds
5299 and where they may be used.
5301 * Added support for having more than one main word list as well as a
5302 special *multi* word list files which will allow multiple word
5303 lists to be treated as one.
5305 * Aspell now uses a completely new word list.
5307 * The apostrophe (') is no longer considered part of the word when
5308 it as at the end of the word such as in `dogs''.
5310 Changes from .31 to .31.1 (June 18, 2000)
5311 =========================================
5313 * Fixed a bug where Aspell would not create a complete dictionary
5314 file on some platforms when the data is 8-bit.
5316 * Added a workaround so Aspell will work with ispell.el 3.3.
5318 * Minor compile fixes so it would compile better with the very latest
5321 * Removed note about compiling in Win32 as I was now able to get it
5324 Changes from .30.1 to .31 (June 11, 2000)
5325 =========================================
5327 * Added support for spell checking run together words.
5329 * Added an option to produce a list of misspelled words from
5332 * More robust error reporting when reading in language data files.
5334 * Fixed a bug that would cause Aspell to crash if the *special* line
5335 was not defined in the language data file.
5337 * Updated Pspell Module.
5341 * Added cross references in "The Aspell Utility Chapter" for easier
5344 Changes from .30 to .30.1 (April 29, 2000)
5345 ==========================================
5347 * Ported Aspell to Win32 platforms.
5349 * Portability fixes which may help Aspell compile on other platforms.
5351 * Aspell will no longer fail if for some reason the mmap fails,
5352 instead it will just read the file in as normal and free the
5355 * Minor changes in the format of the main word list as a result of
5356 the changes, the old format should still work in most cases.
5358 * Fixed a bug where Aspell was ignoring the extension of file names
5359 such as .html or .tex when checking files.
5361 * Fixed a bug where Aspell will go into an infinite loop when
5362 creating the main word list from a word list which has duplicates
5365 * Minor changes to the manual for better clarity.
5367 Changes from .29.1 to .30 (April 2, 2000)
5368 =========================================
5370 * Fixed many of the capitalization bugs found in previous versions of
5373 * Changed the format of the main word list yet again.
5375 * Fixed a bug so that `aspell check' will work on the PowerPC.
5377 * Added ability to change configuration options in the middle of a
5380 * Added words from /usr/dict/words found on most Linux systems as
5381 well as a bunch of commonly used abbreviations to the word list.
5383 * Fixed a bug where Aspell would dump core after reporting certain
5384 errors when compiled with gcc 2.95 or higher. This involved
5385 reworking the Exception heritage to get around a bug in gcc 2.95.
5387 * Added a few more commands to the list of default commands the TeX
5390 * Aspell will now check if a word only contains valid characters
5391 before adding it to any dictionaries. This might mean that you
5392 have to manually delete a few words from your personal word list.
5394 * Added option to ignore case when checking a document.
5396 * Adjusted the parameters of the *normal* suggest mode to so that
5397 significantly less far fetched results are returned in cases such
5398 as tomatoe, which went from 100 suggestions down to 32, at the
5399 expense of getting slightly lower results (less than 1%),
5401 * Improved the edit distance algorithm for slightly faster results.
5403 * Removed the `$$m' command in pipe mode, you should now use `$$cs
5404 mode,MODE' to set the mode and *$$cr mode* to find out the current
5407 * Reworked parts of Aspell to use Pspell services to avoid
5410 * Added a module for the newly released Pspell. It will get
5411 installed with the rest of Aspell.
5413 * Miscellaneous other bug fixes.
5415 Changes from .29 to .29.1 (Feb 18, 2000)
5416 ========================================
5418 * Improved the TeX filter so that it will accept '@' at the beginning
5419 of a command name and ignored trailing '*'s. It also now has
5420 better defaults for which parameters to skip.
5422 * Reworked the main dictionary so that it can be memory mapped in.
5423 This decreases startup time and allows multiple Aspell processes
5424 to use the same memory for the main word list. This also also
5425 made Aspell 64 bit clean so that it should work on an alpha now.
5427 * Fix so that Aspell could compile on platforms that gnu is not yet
5430 * Fixed issue with flock so it would compile on FreeBSD.
5432 * Minor changes in the code to make it more C++ compliant although I
5433 am sure there will still be problems when using some other
5434 compiler other than gcc or egcs.
5436 * Added some comments to the header files to better document a few of
5439 Changes from .28.3 to .29 (Feb 6, 2000)
5440 =======================================
5442 * Fixed a bug in the pipe mode with lines that start with `^$$'.
5444 * Added support for ignoring all words less than or equal to a
5447 * New soundslike code based thanks to the contribution of Björn
5448 Jacke. It now gets all of its data from a table making it easier
5449 for other people to add soundslike code for their native language.
5450 He also converted the metaphone algorithm to table form,
5451 eliminating the need for the old metaphone code.
5453 * Major redesign of the suggestion code for better results.
5455 * Changed the format of the personal word lists. In most cases it
5456 should be converted automatically.
5458 * Changed the format of the main word list.
5460 * Name space cleanup for more consistent naming. I now use name
5461 spaces which means that gcc 2.8.* and egcs 1.0.* will no longer
5464 * Used file locks when reading and saving the personal dictionaries
5465 so that it truly multiprocesses safely.
5467 * Added rudimentary filter support.
5469 * Reworked the configuration system once again. However, the
5470 changes to the end user who does not directly use my library
5473 * Rewrote my code that handles parsing command line parameters so
5474 that it no longer used popt as it was causing too many problems
5475 and didn't integrate well with my new configuration system.
5477 * Fixed pipe mode so that it will properly ignore lines starting with
5478 '~' for better Ispell compatibility.
5480 * Aspell now has a new home page at
5481 `http://aspell.sourceforge.net/'. Please make note of the new URL.
5483 * Miscellaneous manual fixes and clarifications.
5485 Changes from .28.2.1 to .28.3 (Nov 20, 1999)
5486 ============================================
5488 * Fixed a bug that caused Aspell to crash when spell checking words
5489 over 60 characters long.
5491 * Reworked *aspell check* so that
5492 1. You no longer have to hit enter when making a choice.
5494 2. It will now overwrite the original file instead of creating a
5495 new file. An optional backup can be made by using the -b
5498 * Fixed a few bugs in data.cc.
5500 Changes from .28.2 to .28.2.1 (Aug 25, 1999)
5501 ============================================
5503 * Fixed the version number for the shared library.
5505 * Fixed a problem with undefined references when linking to the
5508 Changes from .28.1 to .28.2 (Aug 25, 1999)
5509 ==========================================
5511 * Fixed a bunch of bugs in the language and configuration classes.
5513 * Minor changes in the code so that it could compile with the new gcc
5516 * Changed the output of `dump config' so that default values are
5517 given the value `<default>'. This means that the output can be
5518 used to create a configuration file.
5520 * Added notes on using Aspell with VIM.
5522 Changes from .28 to .28.1 (July 27, 1999)
5523 =========================================
5525 * Removed some debug output
5527 * Changed notes on compiling with gcc 2.8.* as I managed to get it to
5528 compile on my school account
5530 * Avoided including *stdexcept* in `const_string.hh' so that I could
5531 get Aspell to compile on my school account with gcc 2.8.1.
5533 Changes from .27.2 to .28 (July 25, 1999)
5534 =========================================
5536 Provided an iterator for the replacement classes.
5538 * Added support for dumping and creating and merging the personal and
5539 replacement word lists.
5541 * Changed the Aspell utility command line a bit, it now used popt.
5543 * Totally reworked Aspell configuration system. Now Aspell could get
5544 configuration from any of 5 sources: the command line, the
5545 environment variable `ASPELL_CONF', the personal configuration
5546 file, the global configuration file, and finally the compiled-in
5549 * Totally reworked the language class in preparation for my new
5550 language code. See `http://aspell.sourceforge.net/international/'
5551 for more information of what I have in store.
5553 * Added some options to the configure script: -enable-dict-dir=DIR,
5554 -enable-doc-dir=DIR, -enable-debug, and -enable-opt
5556 * Removed some old header files.
5558 * Reorganized the directory structure a bit
5560 * Made the text version of the manual pages slightly easier to read
5562 * Used the `\url' command for urls for better formating of the
5565 Changes from .27.1 to .27.2 (Mar 1, 1999)
5566 =========================================
5568 * Fixed a major bug that caused Aspell to dump core when used without
5571 * Fixed another major bug that caused Aspell to do nothing when used
5572 in interactive mode.
5574 * Added an option to exit in Aspell's interactive mode.
5576 * Removed some old documentation files from the distribution.
5578 * Minor changes to the the section on using Aspell with egcs.
5580 * Minor changes to remove -Wall warnings.
5582 Changes from .27 to .27.1 (Feb 24, 1999)
5583 ========================================
5585 * Fixed a minor compile problem.
5587 * Updated the section on using Aspell with egcs to it. It was now
5588 more clear why the patch was necessary.
5590 Changes from .26.2 to .27 (Feb 22, 1999)
5591 ========================================
5593 * Totally reworked the C++ library which means you may need to change
5594 some things in your code.
5596 * Added support for detachable and multiple personal dictionaries in
5597 the C++ class library.
5599 * The C++ class library now throws exceptions.
5601 * Reworked Aspell ability to learn from users misspellings a bit so
5602 that it now has a memory. For more information see *Note Notes on
5603 Storing Replacement Pairs::.
5605 * Upgraded autoconf to version 2.13 and automake to version 1.4 for
5608 * Fixed the configuration so the `make dist' will work. From now on
5609 Aspell will be distributed with `make dist'.
5611 * Added support to skip over URL's, email addresses and host names.
5613 * Added support for dumping the master and personal word list. You
5614 can now also merge a personal word list. Type aspell -help for
5615 help on using this feature.
5617 * Reorganized the source code.
5619 * Started using proper version numbers for the shared library.
5621 * Fixed a bug that caused Aspell to crash when adding certain
5624 * Fixed the problem with duplicate lines when exiting pipe mode for
5627 Changed from .26.1 to .26.2 (Jan 3, 1998)
5628 =========================================
5630 * Fixed another compile problem. Hopefully this time it will really
5631 compile OK on other peoples machines.
5633 Changed from .26 to .26.1 (Jan 3, 1998)
5634 =======================================
5636 * Fixed a small compile problem in `as_data.cc'.
5638 Changed from .25.1 to .26 (Jan 3, 1999)
5639 =======================================
5641 * Fixed a bug that caused duplicate items to be displayed in the
5642 suggestion list for good. (If it still does it please send me
5645 * Added the ability for Aspell to learn form the users misspellings.
5647 * Library Interface changes. Still more to come ....
5649 * Is now multiprocess safe. When a personal dictionary (or
5650 replacement list) is saved it will now first update the list
5651 against the dictionary on disk in case another process modified it.
5653 * Fixed the bug that caused duplicate output when used non
5654 interactively in pipe mode.
5656 * Dropped support for gcc 2.7.2 as the C++ compiler.
5658 * Updated the How Aspell Works (*Note Aspell Suggestion Strategy::.)
5660 * Added support for the `ASPELL_DATA_DIR' environment variable.
5662 Changes from .25 to .25.1 (Dec 10, 1998)
5663 ========================================
5665 * Fixed the version number so that Aspell reports the correct version
5668 * Changed the note on gcc 2.7.2 compilers to make it clear that only
5669 the C++ compiler cannot be gcc 2.7.2, it is OK if the C compiler
5672 * Updated the TODO list and reorganized it a bit.
5674 * Fixed the directory so that all the documentation will get
5675 installed in ${prefix}/doc/aspell instead of half of it in
5676 ${prefix}/doc/aspell and half of it in ${prefix}/doc/kspell.
5678 Changes from .24 to .25 (Nov 23, 1998)
5679 ======================================
5681 * Total rework of how the main word list is stored. Start up time
5682 decreased to about 1/3 of what it was in .24 and memory usage
5683 decreased to about 2/3. (When used with the provided word list on
5686 Also the format and default locations of the main word list data
5687 files changed in the process and the data is now machine
5688 dependent. The personal word list format, however, stayed the
5691 * Changed the scoring method to produce slightly better results with
5692 words like the vs. teh. And other simpler misspellings where two
5693 letters are swapped.
5695 * Fixed the very unpredictable behavior of the `*', `&', `@'
5696 commands in the pipe mode.
5698 * Added documentations for Aspell pipe mode (also known as `ispell
5699 -a' compatibility mode)
5701 * Added a bunch of Aspell specific extensions to the pipe mode and
5704 * Documented the `to_soundslike' and `soundslike' methods for the
5707 * Changed the scoring method to produce better results for words like
5708 _fone_ vs _phone_ and other words that have a spelling that
5709 doesn't directly relate to how the word sounds by using the phoneme
5710 equivalent of the word in the scoring of it.
5712 * Added the `to_phoneme' and `have_phoneme' methods to the
5713 `SC_Language' class.
5715 * Added the `to_phoneme' method to the `aspell' class.
5717 * Added the framework for being able to learn from the users
5718 misspelling. Right now it just keeps a log of replacements.
5720 * Redid `stl_rope-30.diff'. For some reason the version of patch on
5721 my system refused it.
5723 * Rewrite of the "_Using as a replacement for Ispell_" section and
5724 added the `run-with-aspell' utility as a replacement of the old
5725 method of mapping Ispell to Aspell.
5727 * Fixed a bug that caused duplicate words to appear in the suggestion
5730 Changes from .23 to .24 (Nov 8, 1998)
5731 =====================================
5733 * Fixed my code so that it can once again compile with g++ 2.7.2.
5735 * Rewrote the How It Works chapter.
5737 * Rewrote the Requirement section and added notes on compiling with
5740 * Added a To Do chapter.
5742 * Added a Glossary and References chapter.
5744 * Other minor documentation improvements.
5746 * Internal code documentation improvements.
5748 Changes from .22.1 to .23 (Oct 31, 1998)
5749 ========================================
5751 * Minor documentation fixes.
5753 * Changed the scoring strategy for words with 3 or less letters.
5754 This cut the number of words returned for these roughly in half.
5756 * Expanded the word list to also include *american.0* and
5757 *american.1* from the Ispell distribution. It now includes
5758 *english.0*, *english.1*, *american.0* and *american.1* from the
5759 directory `languages/english' provided with Ispell 3.1.20.
5761 * Added a link to the location of the latest Ispell.el in the
5764 * Started a C interface and added some rough documentation for it.
5766 Changes from .22 to .22.1 (Oct 27, 1998)
5767 ========================================
5769 * Minor bug fixes. I was deleting arrays with delete rather than
5770 delete[]. I was suprised that this had not created a problem.
5772 * Added a simple test program to test for a memory leak present on
5773 some systems. (Only debian slink at the moment.) See the file
5774 memleak-test.cc for more info.
5776 Changes from .21 to .22 (Oct 26, 1998)
5777 ======================================
5779 * Major redesign of the scoring method. It now uses absolute
5780 distances rather than relative scores for more consistent results.
5781 See `suggest.cc' for more info.
5783 * Suggest code rewritten in several places, however the core process
5786 * The `suggest_ultra' method temporarily does nothing. It should be
5787 working again by the next release.
5789 Changes from .20 to .21 (Oct 13, 1998)
5790 ======================================
5792 * Added documentation for aspell::Error
5794 * Changed the library name from `libspell' to `libaspell'. It
5795 should never have been `libspell' in the first place. Sorry for
5796 the incompatibility.
5798 * Added `as_error.hh' to the list of files copied to the include
5799 directory so that you can actually use the library outside of the
5802 * Fixed bug that caused a segmentation fault with words where the
5803 only suggestions was inserting a space or hyphen such as in
5806 * Added the *score* method to `aspell'.
5808 * Changed the scoring method to deal with word when the user uses
5809 "f" in place of "ph" a lot better.
5811 Changes from .11 to .20 (Oct 10, 1998)
5812 ======================================
5814 * _Name change_. Everything that was Kspell is now Aspell. Sorry,
5815 the name Kspell was already used by KDE and I didn't want to cause
5818 * Fixed a bug that causes a segmentation fault when the `HOME'
5819 environment variable doesn't exist.
5821 Changes from .10 to .11 (Sep 12, 1998)
5822 ======================================
5824 * Overhaul of the SC_Language class
5826 * Added documentation for international support
5828 * Added documentation for the C++ library
5830 * Other minor bug fixes.
5833 File: aspell.info, Node: Authors, Next: Copying, Prev: ChangeLog, Up: Top
5838 The following people or companies have contributed a non-trival amount
5839 of code to Aspell and thus own the Copyright to part of Aspell.
5842 Bug fixes and enhancements to `word-list-compress'.
5845 Wrote the Nroff filter.
5848 Enhanced the SGML filter to also support skipping sgml tags such as
5851 Kevin B. Hendricks (and Contributers)
5852 Wrote MySpell which is a simple spell checker library that supports
5853 affix compression. Aspell affix compression code is based on his
5856 Christoph Hintermüller
5857 Added support for loadable filters.
5860 Wrote a locale independent version of strtol and strtod. Wrote
5861 the original loadable filter support however his code has been
5862 completely rewritten by Christoph Hintermüller and Kevin Atkinson.
5865 Wrote the generic soundslike algorithm which gets all of its data
5866 from a file, thus eliminating almost all need for language
5867 specific code from Aspell.
5869 Silicon Graphics Computer Systems, Inc.
5870 Hewlett-Packard Company
5871 Parts of the SGI STL code were used in various places throughout
5875 In addition the authors of some of translated messages did not
5876 release their work into the Public Domain, and thus own the copyright
5877 to the translated text. See the files `*.po' in the `po' directory for
5880 The folowing people also contributed to the development of Aspell
5881 but do not own the Copyright to part of Aspell.
5884 Added gettext support.
5887 Converted the manual to texinfo.
5890 Wrote the original metaphone algorithm; however, he released his
5891 work into the Public Domain.
5894 Converted the metaphone algorithm into C code and made some
5895 enhancements to the original algorithm. He also released his work
5896 into the Public Domain.
5898 Geoff Kuenning (and contributers)
5899 The authors of Ispell. Many of the ideas used in Aspell,
5900 especially with the affix code, were taken from Ispell. However
5901 none of the original Ispell code is used in Aspell.
5905 File: aspell.info, Node: Copying, Prev: Authors, Up: Top
5910 Copyright (C) 2000-2006 Kevin Atkinson.
5912 Permission is granted to copy, distribute and/or modify this document
5913 under the terms of the GNU Free Documentation License, Version 1.1 or
5914 any later version published by the Free Software Foundation; with no
5915 Invariant Sections, no Front-Cover Texts and no Back-Cover Texts. A
5916 copy of the license is included in the section entitled "GNU Free
5917 Documentation License".
5919 The library and utility program is copyright (C) 2000-2006 by Kevin
5920 Atkinson. You can redistribute it and/or modify it under the terms of
5921 the GNU Lesser General Public License as (LGPL) published by the Free
5922 Software Foundation; either version 2.1 of the License, or (at your
5923 option) any later version.
5925 Certain parts of the library, as indicated at the top of the source
5926 file, are under a weaker license. However, all parts of the library
5927 are LGPL Compatible.
5931 * GNU Free Documentation License::
5932 * GNU Lesser General Public License::
5935 File: aspell.info, Node: GNU Free Documentation License, Next: GNU Lesser General Public License, Up: Copying
5937 H.1 GNU Free Documentation License
5938 ==================================
5940 Version 1.2, November 2002
5942 Copyright (C) 2000,2001,2002 Free Software Foundation, Inc.
5943 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA
5945 Everyone is permitted to copy and distribute verbatim copies
5946 of this license document, but changing it is not allowed.
5950 The purpose of this License is to make a manual, textbook, or other
5951 functional and useful document "free" in the sense of freedom: to
5952 assure everyone the effective freedom to copy and redistribute it,
5953 with or without modifying it, either commercially or
5954 noncommercially. Secondarily, this License preserves for the
5955 author and publisher a way to get credit for their work, while not
5956 being considered responsible for modifications made by others.
5958 This License is a kind of "copyleft", which means that derivative
5959 works of the document must themselves be free in the same sense.
5960 It complements the GNU General Public License, which is a copyleft
5961 license designed for free software.
5963 We have designed this License in order to use it for manuals for
5964 free software, because free software needs free documentation: a
5965 free program should come with manuals providing the same freedoms
5966 that the software does. But this License is not limited to
5967 software manuals; it can be used for any textual work, regardless
5968 of subject matter or whether it is published as a printed book.
5969 We recommend this License principally for works whose purpose is
5970 instruction or reference.
5972 1. APPLICABILITY AND DEFINITIONS
5974 This License applies to any manual or other work, in any medium,
5975 that contains a notice placed by the copyright holder saying it
5976 can be distributed under the terms of this License. Such a notice
5977 grants a world-wide, royalty-free license, unlimited in duration,
5978 to use that work under the conditions stated herein. The
5979 "Document", below, refers to any such manual or work. Any member
5980 of the public is a licensee, and is addressed as "you". You
5981 accept the license if you copy, modify or distribute the work in a
5982 way requiring permission under copyright law.
5984 A "Modified Version" of the Document means any work containing the
5985 Document or a portion of it, either copied verbatim, or with
5986 modifications and/or translated into another language.
5988 A "Secondary Section" is a named appendix or a front-matter section
5989 of the Document that deals exclusively with the relationship of the
5990 publishers or authors of the Document to the Document's overall
5991 subject (or to related matters) and contains nothing that could
5992 fall directly within that overall subject. (Thus, if the Document
5993 is in part a textbook of mathematics, a Secondary Section may not
5994 explain any mathematics.) The relationship could be a matter of
5995 historical connection with the subject or with related matters, or
5996 of legal, commercial, philosophical, ethical or political position
5999 The "Invariant Sections" are certain Secondary Sections whose
6000 titles are designated, as being those of Invariant Sections, in
6001 the notice that says that the Document is released under this
6002 License. If a section does not fit the above definition of
6003 Secondary then it is not allowed to be designated as Invariant.
6004 The Document may contain zero Invariant Sections. If the Document
6005 does not identify any Invariant Sections then there are none.
6007 The "Cover Texts" are certain short passages of text that are
6008 listed, as Front-Cover Texts or Back-Cover Texts, in the notice
6009 that says that the Document is released under this License. A
6010 Front-Cover Text may be at most 5 words, and a Back-Cover Text may
6011 be at most 25 words.
6013 A "Transparent" copy of the Document means a machine-readable copy,
6014 represented in a format whose specification is available to the
6015 general public, that is suitable for revising the document
6016 straightforwardly with generic text editors or (for images
6017 composed of pixels) generic paint programs or (for drawings) some
6018 widely available drawing editor, and that is suitable for input to
6019 text formatters or for automatic translation to a variety of
6020 formats suitable for input to text formatters. A copy made in an
6021 otherwise Transparent file format whose markup, or absence of
6022 markup, has been arranged to thwart or discourage subsequent
6023 modification by readers is not Transparent. An image format is
6024 not Transparent if used for any substantial amount of text. A
6025 copy that is not "Transparent" is called "Opaque".
6027 Examples of suitable formats for Transparent copies include plain
6028 ASCII without markup, Texinfo input format, LaTeX input format,
6029 SGML or XML using a publicly available DTD, and
6030 standard-conforming simple HTML, PostScript or PDF designed for
6031 human modification. Examples of transparent image formats include
6032 PNG, XCF and JPG. Opaque formats include proprietary formats that
6033 can be read and edited only by proprietary word processors, SGML or
6034 XML for which the DTD and/or processing tools are not generally
6035 available, and the machine-generated HTML, PostScript or PDF
6036 produced by some word processors for output purposes only.
6038 The "Title Page" means, for a printed book, the title page itself,
6039 plus such following pages as are needed to hold, legibly, the
6040 material this License requires to appear in the title page. For
6041 works in formats which do not have any title page as such, "Title
6042 Page" means the text near the most prominent appearance of the
6043 work's title, preceding the beginning of the body of the text.
6045 A section "Entitled XYZ" means a named subunit of the Document
6046 whose title either is precisely XYZ or contains XYZ in parentheses
6047 following text that translates XYZ in another language. (Here XYZ
6048 stands for a specific section name mentioned below, such as
6049 "Acknowledgements", "Dedications", "Endorsements", or "History".)
6050 To "Preserve the Title" of such a section when you modify the
6051 Document means that it remains a section "Entitled XYZ" according
6054 The Document may include Warranty Disclaimers next to the notice
6055 which states that this License applies to the Document. These
6056 Warranty Disclaimers are considered to be included by reference in
6057 this License, but only as regards disclaiming warranties: any other
6058 implication that these Warranty Disclaimers may have is void and
6059 has no effect on the meaning of this License.
6063 You may copy and distribute the Document in any medium, either
6064 commercially or noncommercially, provided that this License, the
6065 copyright notices, and the license notice saying this License
6066 applies to the Document are reproduced in all copies, and that you
6067 add no other conditions whatsoever to those of this License. You
6068 may not use technical measures to obstruct or control the reading
6069 or further copying of the copies you make or distribute. However,
6070 you may accept compensation in exchange for copies. If you
6071 distribute a large enough number of copies you must also follow
6072 the conditions in section 3.
6074 You may also lend copies, under the same conditions stated above,
6075 and you may publicly display copies.
6077 3. COPYING IN QUANTITY
6079 If you publish printed copies (or copies in media that commonly
6080 have printed covers) of the Document, numbering more than 100, and
6081 the Document's license notice requires Cover Texts, you must
6082 enclose the copies in covers that carry, clearly and legibly, all
6083 these Cover Texts: Front-Cover Texts on the front cover, and
6084 Back-Cover Texts on the back cover. Both covers must also clearly
6085 and legibly identify you as the publisher of these copies. The
6086 front cover must present the full title with all words of the
6087 title equally prominent and visible. You may add other material
6088 on the covers in addition. Copying with changes limited to the
6089 covers, as long as they preserve the title of the Document and
6090 satisfy these conditions, can be treated as verbatim copying in
6093 If the required texts for either cover are too voluminous to fit
6094 legibly, you should put the first ones listed (as many as fit
6095 reasonably) on the actual cover, and continue the rest onto
6098 If you publish or distribute Opaque copies of the Document
6099 numbering more than 100, you must either include a
6100 machine-readable Transparent copy along with each Opaque copy, or
6101 state in or with each Opaque copy a computer-network location from
6102 which the general network-using public has access to download
6103 using public-standard network protocols a complete Transparent
6104 copy of the Document, free of added material. If you use the
6105 latter option, you must take reasonably prudent steps, when you
6106 begin distribution of Opaque copies in quantity, to ensure that
6107 this Transparent copy will remain thus accessible at the stated
6108 location until at least one year after the last time you
6109 distribute an Opaque copy (directly or through your agents or
6110 retailers) of that edition to the public.
6112 It is requested, but not required, that you contact the authors of
6113 the Document well before redistributing any large number of
6114 copies, to give them a chance to provide you with an updated
6115 version of the Document.
6119 You may copy and distribute a Modified Version of the Document
6120 under the conditions of sections 2 and 3 above, provided that you
6121 release the Modified Version under precisely this License, with
6122 the Modified Version filling the role of the Document, thus
6123 licensing distribution and modification of the Modified Version to
6124 whoever possesses a copy of it. In addition, you must do these
6125 things in the Modified Version:
6127 A. Use in the Title Page (and on the covers, if any) a title
6128 distinct from that of the Document, and from those of
6129 previous versions (which should, if there were any, be listed
6130 in the History section of the Document). You may use the
6131 same title as a previous version if the original publisher of
6132 that version gives permission.
6134 B. List on the Title Page, as authors, one or more persons or
6135 entities responsible for authorship of the modifications in
6136 the Modified Version, together with at least five of the
6137 principal authors of the Document (all of its principal
6138 authors, if it has fewer than five), unless they release you
6139 from this requirement.
6141 C. State on the Title page the name of the publisher of the
6142 Modified Version, as the publisher.
6144 D. Preserve all the copyright notices of the Document.
6146 E. Add an appropriate copyright notice for your modifications
6147 adjacent to the other copyright notices.
6149 F. Include, immediately after the copyright notices, a license
6150 notice giving the public permission to use the Modified
6151 Version under the terms of this License, in the form shown in
6154 G. Preserve in that license notice the full lists of Invariant
6155 Sections and required Cover Texts given in the Document's
6158 H. Include an unaltered copy of this License.
6160 I. Preserve the section Entitled "History", Preserve its Title,
6161 and add to it an item stating at least the title, year, new
6162 authors, and publisher of the Modified Version as given on
6163 the Title Page. If there is no section Entitled "History" in
6164 the Document, create one stating the title, year, authors,
6165 and publisher of the Document as given on its Title Page,
6166 then add an item describing the Modified Version as stated in
6167 the previous sentence.
6169 J. Preserve the network location, if any, given in the Document
6170 for public access to a Transparent copy of the Document, and
6171 likewise the network locations given in the Document for
6172 previous versions it was based on. These may be placed in
6173 the "History" section. You may omit a network location for a
6174 work that was published at least four years before the
6175 Document itself, or if the original publisher of the version
6176 it refers to gives permission.
6178 K. For any section Entitled "Acknowledgements" or "Dedications",
6179 Preserve the Title of the section, and preserve in the
6180 section all the substance and tone of each of the contributor
6181 acknowledgements and/or dedications given therein.
6183 L. Preserve all the Invariant Sections of the Document,
6184 unaltered in their text and in their titles. Section numbers
6185 or the equivalent are not considered part of the section
6188 M. Delete any section Entitled "Endorsements". Such a section
6189 may not be included in the Modified Version.
6191 N. Do not retitle any existing section to be Entitled
6192 "Endorsements" or to conflict in title with any Invariant
6195 O. Preserve any Warranty Disclaimers.
6197 If the Modified Version includes new front-matter sections or
6198 appendices that qualify as Secondary Sections and contain no
6199 material copied from the Document, you may at your option
6200 designate some or all of these sections as invariant. To do this,
6201 add their titles to the list of Invariant Sections in the Modified
6202 Version's license notice. These titles must be distinct from any
6203 other section titles.
6205 You may add a section Entitled "Endorsements", provided it contains
6206 nothing but endorsements of your Modified Version by various
6207 parties--for example, statements of peer review or that the text
6208 has been approved by an organization as the authoritative
6209 definition of a standard.
6211 You may add a passage of up to five words as a Front-Cover Text,
6212 and a passage of up to 25 words as a Back-Cover Text, to the end
6213 of the list of Cover Texts in the Modified Version. Only one
6214 passage of Front-Cover Text and one of Back-Cover Text may be
6215 added by (or through arrangements made by) any one entity. If the
6216 Document already includes a cover text for the same cover,
6217 previously added by you or by arrangement made by the same entity
6218 you are acting on behalf of, you may not add another; but you may
6219 replace the old one, on explicit permission from the previous
6220 publisher that added the old one.
6222 The author(s) and publisher(s) of the Document do not by this
6223 License give permission to use their names for publicity for or to
6224 assert or imply endorsement of any Modified Version.
6226 5. COMBINING DOCUMENTS
6228 You may combine the Document with other documents released under
6229 this License, under the terms defined in section 4 above for
6230 modified versions, provided that you include in the combination
6231 all of the Invariant Sections of all of the original documents,
6232 unmodified, and list them all as Invariant Sections of your
6233 combined work in its license notice, and that you preserve all
6234 their Warranty Disclaimers.
6236 The combined work need only contain one copy of this License, and
6237 multiple identical Invariant Sections may be replaced with a single
6238 copy. If there are multiple Invariant Sections with the same name
6239 but different contents, make the title of each such section unique
6240 by adding at the end of it, in parentheses, the name of the
6241 original author or publisher of that section if known, or else a
6242 unique number. Make the same adjustment to the section titles in
6243 the list of Invariant Sections in the license notice of the
6246 In the combination, you must combine any sections Entitled
6247 "History" in the various original documents, forming one section
6248 Entitled "History"; likewise combine any sections Entitled
6249 "Acknowledgements", and any sections Entitled "Dedications". You
6250 must delete all sections Entitled "Endorsements."
6252 6. COLLECTIONS OF DOCUMENTS
6254 You may make a collection consisting of the Document and other
6255 documents released under this License, and replace the individual
6256 copies of this License in the various documents with a single copy
6257 that is included in the collection, provided that you follow the
6258 rules of this License for verbatim copying of each of the
6259 documents in all other respects.
6261 You may extract a single document from such a collection, and
6262 distribute it individually under this License, provided you insert
6263 a copy of this License into the extracted document, and follow
6264 this License in all other respects regarding verbatim copying of
6267 7. AGGREGATION WITH INDEPENDENT WORKS
6269 A compilation of the Document or its derivatives with other
6270 separate and independent documents or works, in or on a volume of
6271 a storage or distribution medium, is called an "aggregate" if the
6272 copyright resulting from the compilation is not used to limit the
6273 legal rights of the compilation's users beyond what the individual
6274 works permit. When the Document is included in an aggregate, this
6275 License does not apply to the other works in the aggregate which
6276 are not themselves derivative works of the Document.
6278 If the Cover Text requirement of section 3 is applicable to these
6279 copies of the Document, then if the Document is less than one half
6280 of the entire aggregate, the Document's Cover Texts may be placed
6281 on covers that bracket the Document within the aggregate, or the
6282 electronic equivalent of covers if the Document is in electronic
6283 form. Otherwise they must appear on printed covers that bracket
6284 the whole aggregate.
6288 Translation is considered a kind of modification, so you may
6289 distribute translations of the Document under the terms of section
6290 4. Replacing Invariant Sections with translations requires special
6291 permission from their copyright holders, but you may include
6292 translations of some or all Invariant Sections in addition to the
6293 original versions of these Invariant Sections. You may include a
6294 translation of this License, and all the license notices in the
6295 Document, and any Warranty Disclaimers, provided that you also
6296 include the original English version of this License and the
6297 original versions of those notices and disclaimers. In case of a
6298 disagreement between the translation and the original version of
6299 this License or a notice or disclaimer, the original version will
6302 If a section in the Document is Entitled "Acknowledgements",
6303 "Dedications", or "History", the requirement (section 4) to
6304 Preserve its Title (section 1) will typically require changing the
6309 You may not copy, modify, sublicense, or distribute the Document
6310 except as expressly provided for under this License. Any other
6311 attempt to copy, modify, sublicense or distribute the Document is
6312 void, and will automatically terminate your rights under this
6313 License. However, parties who have received copies, or rights,
6314 from you under this License will not have their licenses
6315 terminated so long as such parties remain in full compliance.
6317 10. FUTURE REVISIONS OF THIS LICENSE
6319 The Free Software Foundation may publish new, revised versions of
6320 the GNU Free Documentation License from time to time. Such new
6321 versions will be similar in spirit to the present version, but may
6322 differ in detail to address new problems or concerns. See
6323 `http://www.gnu.org/copyleft/'.
6325 Each version of the License is given a distinguishing version
6326 number. If the Document specifies that a particular numbered
6327 version of this License "or any later version" applies to it, you
6328 have the option of following the terms and conditions either of
6329 that specified version or of any later version that has been
6330 published (not as a draft) by the Free Software Foundation. If
6331 the Document does not specify a version number of this License,
6332 you may choose any version ever published (not as a draft) by the
6333 Free Software Foundation.
6335 H.1.1 ADDENDUM: How to use this License for your documents
6336 ----------------------------------------------------------
6338 To use this License in a document you have written, include a copy of
6339 the License in the document and put the following copyright and license
6340 notices just after the title page:
6342 Copyright (C) YEAR YOUR NAME.
6343 Permission is granted to copy, distribute and/or modify this document
6344 under the terms of the GNU Free Documentation License, Version 1.2
6345 or any later version published by the Free Software Foundation;
6346 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
6347 Texts. A copy of the license is included in the section entitled ``GNU
6348 Free Documentation License''.
6350 If you have Invariant Sections, Front-Cover Texts and Back-Cover
6351 Texts, replace the "with...Texts." line with this:
6353 with the Invariant Sections being LIST THEIR TITLES, with
6354 the Front-Cover Texts being LIST, and with the Back-Cover Texts
6357 If you have Invariant Sections without Cover Texts, or some other
6358 combination of the three, merge those two alternatives to suit the
6361 If your document contains nontrivial examples of program code, we
6362 recommend releasing these examples in parallel under your choice of
6363 free software license, such as the GNU General Public License, to
6364 permit their use in free software.
6367 File: aspell.info, Node: GNU Lesser General Public License, Prev: GNU Free Documentation License, Up: Copying
6369 H.2 GNU Lesser General Public License
6370 =====================================
6372 Version 2.1, February 1999
6374 Copyright (C) 1991, 1999 Free Software Foundation, Inc.
6375 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA
6377 Everyone is permitted to copy and distribute verbatim copies
6378 of this license document, but changing it is not allowed.
6380 [This is the first released version of the Lesser GPL. It also counts
6381 as the successor of the GNU Library Public License, version 2, hence the
6382 version number 2.1.]
6387 The licenses for most software are designed to take away your freedom
6388 to share and change it. By contrast, the GNU General Public Licenses
6389 are intended to guarantee your freedom to share and change free
6390 software--to make sure the software is free for all its users.
6392 This license, the Lesser General Public License, applies to some
6393 specially designated software--typically libraries--of the Free
6394 Software Foundation and other authors who decide to use it. You can use
6395 it too, but we suggest you first think carefully about whether this
6396 license or the ordinary General Public License is the better strategy to
6397 use in any particular case, based on the explanations below.
6399 When we speak of free software, we are referring to freedom of use,
6400 not price. Our General Public Licenses are designed to make sure that
6401 you have the freedom to distribute copies of free software (and charge
6402 for this service if you wish); that you receive source code or can get
6403 it if you want it; that you can change the software and use pieces of it
6404 in new free programs; and that you are informed that you can do these
6407 To protect your rights, we need to make restrictions that forbid
6408 distributors to deny you these rights or to ask you to surrender these
6409 rights. These restrictions translate to certain responsibilities for
6410 you if you distribute copies of the library or if you modify it.
6412 For example, if you distribute copies of the library, whether gratis
6413 or for a fee, you must give the recipients all the rights that we gave
6414 you. You must make sure that they, too, receive or can get the source
6415 code. If you link other code with the library, you must provide
6416 complete object files to the recipients, so that they can relink them
6417 with the library after making changes to the library and recompiling
6418 it. And you must show them these terms so they know their rights.
6420 We protect your rights with a two-step method: (1) we copyright the
6421 library, and (2) we offer you this license, which gives you legal
6422 permission to copy, distribute and/or modify the library.
6424 To protect each distributor, we want to make it very clear that
6425 there is no warranty for the free library. Also, if the library is
6426 modified by someone else and passed on, the recipients should know that
6427 what they have is not the original version, so that the original
6428 author's reputation will not be affected by problems that might be
6429 introduced by others.
6431 Finally, software patents pose a constant threat to the existence of
6432 any free program. We wish to make sure that a company cannot
6433 effectively restrict the users of a free program by obtaining a
6434 restrictive license from a patent holder. Therefore, we insist that
6435 any patent license obtained for a version of the library must be
6436 consistent with the full freedom of use specified in this license.
6438 Most GNU software, including some libraries, is covered by the
6439 ordinary GNU General Public License. This license, the GNU Lesser
6440 General Public License, applies to certain designated libraries, and is
6441 quite different from the ordinary General Public License. We use this
6442 license for certain libraries in order to permit linking those
6443 libraries into non-free programs.
6445 When a program is linked with a library, whether statically or using
6446 a shared library, the combination of the two is legally speaking a
6447 combined work, a derivative of the original library. The ordinary
6448 General Public License therefore permits such linking only if the
6449 entire combination fits its criteria of freedom. The Lesser General
6450 Public License permits more lax criteria for linking other code with
6453 We call this license the "Lesser" General Public License because it
6454 does _Less_ to protect the user's freedom than the ordinary General
6455 Public License. It also provides other free software developers Less
6456 of an advantage over competing non-free programs. These disadvantages
6457 are the reason we use the ordinary General Public License for many
6458 libraries. However, the Lesser license provides advantages in certain
6459 special circumstances.
6461 For example, on rare occasions, there may be a special need to
6462 encourage the widest possible use of a certain library, so that it
6463 becomes a de-facto standard. To achieve this, non-free programs must be
6464 allowed to use the library. A more frequent case is that a free
6465 library does the same job as widely used non-free libraries. In this
6466 case, there is little to gain by limiting the free library to free
6467 software only, so we use the Lesser General Public License.
6469 In other cases, permission to use a particular library in non-free
6470 programs enables a greater number of people to use a large body of free
6471 software. For example, permission to use the GNU C Library in non-free
6472 programs enables many more people to use the whole GNU operating
6473 system, as well as its variant, the GNU/Linux operating system.
6475 Although the Lesser General Public License is Less protective of the
6476 users' freedom, it does ensure that the user of a program that is
6477 linked with the Library has the freedom and the wherewithal to run that
6478 program using a modified version of the Library.
6480 The precise terms and conditions for copying, distribution and
6481 modification follow. Pay close attention to the difference between a
6482 "work based on the library" and a "work that uses the library". The
6483 former contains code derived from the library, whereas the latter must
6484 be combined with the library in order to run.
6486 GNU LESSER GENERAL PUBLIC LICENSE
6487 TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
6488 0. This License Agreement applies to any software library or other
6489 program which contains a notice placed by the copyright holder or
6490 other authorized party saying it may be distributed under the
6491 terms of this Lesser General Public License (also called "this
6492 License"). Each licensee is addressed as "you".
6494 A "library" means a collection of software functions and/or data
6495 prepared so as to be conveniently linked with application programs
6496 (which use some of those functions and data) to form executables.
6498 The "Library", below, refers to any such software library or work
6499 which has been distributed under these terms. A "work based on the
6500 Library" means either the Library or any derivative work under
6501 copyright law: that is to say, a work containing the Library or a
6502 portion of it, either verbatim or with modifications and/or
6503 translated straightforwardly into another language. (Hereinafter,
6504 translation is included without limitation in the term
6507 "Source code" for a work means the preferred form of the work for
6508 making modifications to it. For a library, complete source code
6509 means all the source code for all modules it contains, plus any
6510 associated interface definition files, plus the scripts used to
6511 control compilation and installation of the library.
6513 Activities other than copying, distribution and modification are
6514 not covered by this License; they are outside its scope. The act
6515 of running a program using the Library is not restricted, and
6516 output from such a program is covered only if its contents
6517 constitute a work based on the Library (independent of the use of
6518 the Library in a tool for writing it). Whether that is true
6519 depends on what the Library does and what the program that uses
6522 1. You may copy and distribute verbatim copies of the Library's
6523 complete source code as you receive it, in any medium, provided
6524 that you conspicuously and appropriately publish on each copy an
6525 appropriate copyright notice and disclaimer of warranty; keep
6526 intact all the notices that refer to this License and to the
6527 absence of any warranty; and distribute a copy of this License
6528 along with the Library.
6530 You may charge a fee for the physical act of transferring a copy,
6531 and you may at your option offer warranty protection in exchange
6534 2. You may modify your copy or copies of the Library or any portion
6535 of it, thus forming a work based on the Library, and copy and
6536 distribute such modifications or work under the terms of Section 1
6537 above, provided that you also meet all of these conditions:
6539 a. The modified work must itself be a software library.
6541 b. You must cause the files modified to carry prominent notices
6542 stating that you changed the files and the date of any change.
6544 c. You must cause the whole of the work to be licensed at no
6545 charge to all third parties under the terms of this License.
6547 d. If a facility in the modified Library refers to a function or
6548 a table of data to be supplied by an application program that
6549 uses the facility, other than as an argument passed when the
6550 facility is invoked, then you must make a good faith effort
6551 to ensure that, in the event an application does not supply
6552 such function or table, the facility still operates, and
6553 performs whatever part of its purpose remains meaningful.
6555 (For example, a function in a library to compute square roots
6556 has a purpose that is entirely well-defined independent of the
6557 application. Therefore, Subsection 2d requires that any
6558 application-supplied function or table used by this function
6559 must be optional: if the application does not supply it, the
6560 square root function must still compute square roots.)
6562 These requirements apply to the modified work as a whole. If
6563 identifiable sections of that work are not derived from the
6564 Library, and can be reasonably considered independent and separate
6565 works in themselves, then this License, and its terms, do not
6566 apply to those sections when you distribute them as separate
6567 works. But when you distribute the same sections as part of a
6568 whole which is a work based on the Library, the distribution of
6569 the whole must be on the terms of this License, whose permissions
6570 for other licensees extend to the entire whole, and thus to each
6571 and every part regardless of who wrote it.
6573 Thus, it is not the intent of this section to claim rights or
6574 contest your rights to work written entirely by you; rather, the
6575 intent is to exercise the right to control the distribution of
6576 derivative or collective works based on the Library.
6578 In addition, mere aggregation of another work not based on the
6579 Library with the Library (or with a work based on the Library) on
6580 a volume of a storage or distribution medium does not bring the
6581 other work under the scope of this License.
6583 3. You may opt to apply the terms of the ordinary GNU General Public
6584 License instead of this License to a given copy of the Library.
6585 To do this, you must alter all the notices that refer to this
6586 License, so that they refer to the ordinary GNU General Public
6587 License, version 2, instead of to this License. (If a newer
6588 version than version 2 of the ordinary GNU General Public License
6589 has appeared, then you can specify that version instead if you
6590 wish.) Do not make any other change in these notices.
6592 Once this change is made in a given copy, it is irreversible for
6593 that copy, so the ordinary GNU General Public License applies to
6594 all subsequent copies and derivative works made from that copy.
6596 This option is useful when you wish to copy part of the code of
6597 the Library into a program that is not a library.
6599 4. You may copy and distribute the Library (or a portion or
6600 derivative of it, under Section 2) in object code or executable
6601 form under the terms of Sections 1 and 2 above provided that you
6602 accompany it with the complete corresponding machine-readable
6603 source code, which must be distributed under the terms of Sections
6604 1 and 2 above on a medium customarily used for software
6607 If distribution of object code is made by offering access to copy
6608 from a designated place, then offering equivalent access to copy
6609 the source code from the same place satisfies the requirement to
6610 distribute the source code, even though third parties are not
6611 compelled to copy the source along with the object code.
6613 5. A program that contains no derivative of any portion of the
6614 Library, but is designed to work with the Library by being
6615 compiled or linked with it, is called a "work that uses the
6616 Library". Such a work, in isolation, is not a derivative work of
6617 the Library, and therefore falls outside the scope of this License.
6619 However, linking a "work that uses the Library" with the Library
6620 creates an executable that is a derivative of the Library (because
6621 it contains portions of the Library), rather than a "work that
6622 uses the library". The executable is therefore covered by this
6623 License. Section 6 states terms for distribution of such
6626 When a "work that uses the Library" uses material from a header
6627 file that is part of the Library, the object code for the work may
6628 be a derivative work of the Library even though the source code is
6629 not. Whether this is true is especially significant if the work
6630 can be linked without the Library, or if the work is itself a
6631 library. The threshold for this to be true is not precisely
6634 If such an object file uses only numerical parameters, data
6635 structure layouts and accessors, and small macros and small inline
6636 functions (ten lines or less in length), then the use of the object
6637 file is unrestricted, regardless of whether it is legally a
6638 derivative work. (Executables containing this object code plus
6639 portions of the Library will still fall under Section 6.)
6641 Otherwise, if the work is a derivative of the Library, you may
6642 distribute the object code for the work under the terms of Section
6643 6. Any executables containing that work also fall under Section 6,
6644 whether or not they are linked directly with the Library itself.
6646 6. As an exception to the Sections above, you may also combine or
6647 link a "work that uses the Library" with the Library to produce a
6648 work containing portions of the Library, and distribute that work
6649 under terms of your choice, provided that the terms permit
6650 modification of the work for the customer's own use and reverse
6651 engineering for debugging such modifications.
6653 You must give prominent notice with each copy of the work that the
6654 Library is used in it and that the Library and its use are covered
6655 by this License. You must supply a copy of this License. If the
6656 work during execution displays copyright notices, you must include
6657 the copyright notice for the Library among them, as well as a
6658 reference directing the user to the copy of this License. Also,
6659 you must do one of these things:
6661 a. Accompany the work with the complete corresponding
6662 machine-readable source code for the Library including
6663 whatever changes were used in the work (which must be
6664 distributed under Sections 1 and 2 above); and, if the work
6665 is an executable linked with the Library, with the complete
6666 machine-readable "work that uses the Library", as object code
6667 and/or source code, so that the user can modify the Library
6668 and then relink to produce a modified executable containing
6669 the modified Library. (It is understood that the user who
6670 changes the contents of definitions files in the Library will
6671 not necessarily be able to recompile the application to use
6672 the modified definitions.)
6674 b. Use a suitable shared library mechanism for linking with the
6675 Library. A suitable mechanism is one that (1) uses at run
6676 time a copy of the library already present on the user's
6677 computer system, rather than copying library functions into
6678 the executable, and (2) will operate properly with a modified
6679 version of the library, if the user installs one, as long as
6680 the modified version is interface-compatible with the version
6681 that the work was made with.
6683 c. Accompany the work with a written offer, valid for at least
6684 three years, to give the same user the materials specified in
6685 Subsection 6a, above, for a charge no more than the cost of
6686 performing this distribution.
6688 d. If distribution of the work is made by offering access to copy
6689 from a designated place, offer equivalent access to copy the
6690 above specified materials from the same place.
6692 e. Verify that the user has already received a copy of these
6693 materials or that you have already sent this user a copy.
6695 For an executable, the required form of the "work that uses the
6696 Library" must include any data and utility programs needed for
6697 reproducing the executable from it. However, as a special
6698 exception, the materials to be distributed need not include
6699 anything that is normally distributed (in either source or binary
6700 form) with the major components (compiler, kernel, and so on) of
6701 the operating system on which the executable runs, unless that
6702 component itself accompanies the executable.
6704 It may happen that this requirement contradicts the license
6705 restrictions of other proprietary libraries that do not normally
6706 accompany the operating system. Such a contradiction means you
6707 cannot use both them and the Library together in an executable
6708 that you distribute.
6710 7. You may place library facilities that are a work based on the
6711 Library side-by-side in a single library together with other
6712 library facilities not covered by this License, and distribute
6713 such a combined library, provided that the separate distribution
6714 of the work based on the Library and of the other library
6715 facilities is otherwise permitted, and provided that you do these
6718 a. Accompany the combined library with a copy of the same work
6719 based on the Library, uncombined with any other library
6720 facilities. This must be distributed under the terms of the
6723 b. Give prominent notice with the combined library of the fact
6724 that part of it is a work based on the Library, and explaining
6725 where to find the accompanying uncombined form of the same
6728 8. You may not copy, modify, sublicense, link with, or distribute the
6729 Library except as expressly provided under this License. Any
6730 attempt otherwise to copy, modify, sublicense, link with, or
6731 distribute the Library is void, and will automatically terminate
6732 your rights under this License. However, parties who have
6733 received copies, or rights, from you under this License will not
6734 have their licenses terminated so long as such parties remain in
6737 9. You are not required to accept this License, since you have not
6738 signed it. However, nothing else grants you permission to modify
6739 or distribute the Library or its derivative works. These actions
6740 are prohibited by law if you do not accept this License.
6741 Therefore, by modifying or distributing the Library (or any work
6742 based on the Library), you indicate your acceptance of this
6743 License to do so, and all its terms and conditions for copying,
6744 distributing or modifying the Library or works based on it.
6746 10. Each time you redistribute the Library (or any work based on the
6747 Library), the recipient automatically receives a license from the
6748 original licensor to copy, distribute, link with or modify the
6749 Library subject to these terms and conditions. You may not impose
6750 any further restrictions on the recipients' exercise of the rights
6751 granted herein. You are not responsible for enforcing compliance
6752 by third parties with this License.
6754 11. If, as a consequence of a court judgment or allegation of patent
6755 infringement or for any other reason (not limited to patent
6756 issues), conditions are imposed on you (whether by court order,
6757 agreement or otherwise) that contradict the conditions of this
6758 License, they do not excuse you from the conditions of this
6759 License. If you cannot distribute so as to satisfy simultaneously
6760 your obligations under this License and any other pertinent
6761 obligations, then as a consequence you may not distribute the
6762 Library at all. For example, if a patent license would not permit
6763 royalty-free redistribution of the Library by all those who
6764 receive copies directly or indirectly through you, then the only
6765 way you could satisfy both it and this License would be to refrain
6766 entirely from distribution of the Library.
6768 If any portion of this section is held invalid or unenforceable
6769 under any particular circumstance, the balance of the section is
6770 intended to apply, and the section as a whole is intended to apply
6771 in other circumstances.
6773 It is not the purpose of this section to induce you to infringe any
6774 patents or other property right claims or to contest validity of
6775 any such claims; this section has the sole purpose of protecting
6776 the integrity of the free software distribution system which is
6777 implemented by public license practices. Many people have made
6778 generous contributions to the wide range of software distributed
6779 through that system in reliance on consistent application of that
6780 system; it is up to the author/donor to decide if he or she is
6781 willing to distribute software through any other system and a
6782 licensee cannot impose that choice.
6784 This section is intended to make thoroughly clear what is believed
6785 to be a consequence of the rest of this License.
6787 12. If the distribution and/or use of the Library is restricted in
6788 certain countries either by patents or by copyrighted interfaces,
6789 the original copyright holder who places the Library under this
6790 License may add an explicit geographical distribution limitation
6791 excluding those countries, so that distribution is permitted only
6792 in or among countries not thus excluded. In such case, this
6793 License incorporates the limitation as if written in the body of
6796 13. The Free Software Foundation may publish revised and/or new
6797 versions of the Lesser General Public License from time to time.
6798 Such new versions will be similar in spirit to the present version,
6799 but may differ in detail to address new problems or concerns.
6801 Each version is given a distinguishing version number. If the
6802 Library specifies a version number of this License which applies
6803 to it and "any later version", you have the option of following
6804 the terms and conditions either of that version or of any later
6805 version published by the Free Software Foundation. If the Library
6806 does not specify a license version number, you may choose any
6807 version ever published by the Free Software Foundation.
6809 14. If you wish to incorporate parts of the Library into other free
6810 programs whose distribution conditions are incompatible with these,
6811 write to the author to ask for permission. For software which is
6812 copyrighted by the Free Software Foundation, write to the Free
6813 Software Foundation; we sometimes make exceptions for this. Our
6814 decision will be guided by the two goals of preserving the free
6815 status of all derivatives of our free software and of promoting
6816 the sharing and reuse of software generally.
6819 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
6820 WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE
6821 LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
6822 HOLDERS AND/OR OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT
6823 WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT
6824 NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
6825 FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE
6826 QUALITY AND PERFORMANCE OF THE LIBRARY IS WITH YOU. SHOULD THE
6827 LIBRARY PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY
6828 SERVICING, REPAIR OR CORRECTION.
6830 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
6831 WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY
6832 MODIFY AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE
6833 LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL,
6834 INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR
6835 INABILITY TO USE THE LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF
6836 DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU
6837 OR THIRD PARTIES OR A FAILURE OF THE LIBRARY TO OPERATE WITH ANY
6838 OTHER SOFTWARE), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN
6839 ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
6841 END OF TERMS AND CONDITIONS
6842 H.2.2 How to Apply These Terms to Your New Libraries
6843 ----------------------------------------------------
6845 If you develop a new library, and you want it to be of the greatest
6846 possible use to the public, we recommend making it free software that
6847 everyone can redistribute and change. You can do so by permitting
6848 redistribution under these terms (or, alternatively, under the terms of
6849 the ordinary General Public License).
6851 To apply these terms, attach the following notices to the library.
6852 It is safest to attach them to the start of each source file to most
6853 effectively convey the exclusion of warranty; and each file should have
6854 at least the "copyright" line and a pointer to where the full notice is
6857 ONE LINE TO GIVE THE LIBRARY'S NAME AND AN IDEA OF WHAT IT DOES.
6858 Copyright (C) YEAR NAME OF AUTHOR
6860 This library is free software; you can redistribute it and/or modify it
6861 under the terms of the GNU Lesser General Public License as published by
6862 the Free Software Foundation; either version 2.1 of the License, or (at
6863 your option) any later version.
6865 This library is distributed in the hope that it will be useful, but
6866 WITHOUT ANY WARRANTY; without even the implied warranty of
6867 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
6868 Lesser General Public License for more details.
6870 You should have received a copy of the GNU Lesser General Public
6871 License along with this library; if not, write to the Free Software
6872 Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307,
6875 Also add information on how to contact you by electronic and paper
6878 You should also get your employer (if you work as a programmer) or
6879 your school, if any, to sign a "copyright disclaimer" for the library,
6880 if necessary. Here is a sample; alter the names:
6882 Yoyodyne, Inc., hereby disclaims all copyright interest in the library
6883 `Frob' (a library for tweaking knobs) written by James Random Hacker.
6885 SIGNATURE OF TY COON, 1 April 1990
6886 Ty Coon, President of Vice
6888 That's all there is to it!
6899 Node: Introduction
\7f3602
6900 Node: Support
\7f7428
6901 Node: Basic Usage
\7f7898
6902 Node: Spellchecking Individual Files
\7f8230
6903 Node: Using Aspell as a Replacement for Ispell
\7f11546
6904 Ref: Differences From Ispell
\7f13496
6905 Node: Using Aspell with other Applications
\7f17114
6906 Node: Customizing Aspell
\7f19498
6907 Node: Specifying Options
\7f20776
6908 Node: The Options
\7f26863
6909 Node: Dumping Configuration Values
\7f38280
6910 Node: Notes on Various Options
\7f38917
6911 Node: Notes on Various Filters and Filter Modes
\7f39250
6912 Node: Notes on the Prefix Option
\7f48179
6913 Node: Notes on Typo-Analysis
\7f49034
6914 Node: Notes on the Different Suggestion Modes
\7f51142
6915 Node: Working With Dictionaries
\7f53458
6916 Node: Using aspell-import
\7f54007
6917 Node: How Aspell Selects an Appropriate Dictionary
\7f54832
6918 Node: Listing Available Dictionaries
\7f56946
6919 Node: Dumping the Contents of the Word List
\7f57391
6920 Node: Creating an Individual Word List
\7f58004
6921 Node: Working With Affix Info in Word Lists
\7f61229
6922 Node: Format of the Personal and Replacement Dictionaries
\7f64908
6923 Ref: Format of the Personal Dictionary
\7f65213
6924 Node: Using Multi Dictionaries
\7f66615
6925 Node: Dictionary Naming
\7f67503
6926 Node: AWLI files
\7f68769
6927 Node: Writing programs to use Aspell
\7f69563
6928 Node: Through the C API
\7f70105
6929 Node: Through A Pipe
\7f77834
6930 Node: Notes on Storing Replacement Pairs
\7f82893
6931 Node: Adding Support For Other Languages
\7f84279
6932 Node: The Language Data File
\7f85223
6933 Ref: data-encoding
\7f86634
6934 Node: Compiling the Word List
\7f90884
6935 Node: Phonetic Code
\7f91339
6936 Node: The Simple Soundslike
\7f107350
6937 Node: Replacement Tables
\7f108584
6938 Node: Affix Compression
\7f109568
6939 Node: Controlling the Behavior of Run-together Words
\7f114865
6940 Node: Creating A New Character Set
\7f115793
6941 Node: Creating An Official Dictionary Package
\7f117028
6942 Node: Implementation Notes
\7f117760
6943 Node: Aspell Suggestion Strategy
\7f118042
6944 Node: Notes on 8-bit Characters
\7f120474
6945 Node: Languages Which Aspell can Support
\7f124322
6946 Node: Supported
\7f124924
6947 Node: Unsupported
\7f141865
6948 Node: Multiple Scripts
\7f145692
6949 Node: Planned Dictionaries
\7f146530
6950 Node: References
\7f150615
6951 Node: Language Related Issues
\7f151634
6952 Node: Compound Words
\7f152179
6953 Node: Words With Symbols in Them
\7f155124
6954 Node: Unicode Normalization
\7f158400
6955 Node: German Sharp S
\7f159350
6956 Node: Context Sensitive Spelling
\7f160014
6957 Node: To Do
\7f160602
6958 Node: Important Items
\7f160806
6959 Node: Other Items
\7f165096
6960 Node: Notes on Various Items
\7f167154
6961 Node: Word skipping by context
\7f167390
6962 Node: Hidden Markov Model
\7f169262
6963 Node: Email the Personal Dictionary
\7f170125
6964 Node: Installing
\7f171189
6965 Node: Generic Install Instructions
\7f172220
6966 Node: HTML Manuals and "make clean"
\7f173848
6967 Node: Curses Notes
\7f174391
6968 Node: Loadable Filter Notes
\7f176341
6969 Node: Upgrading from Aspell 0.50
\7f176980
6970 Ref: Binary Compatibility
\7f178087
6971 Node: Upgrading from Aspell .33/Pspell .12
\7f179638
6972 Node: Upgrading from a Pre-0.50 snapshot
\7f182150
6973 Node: WIN32 Notes
\7f182662
6974 Ref: Win32-Directories
\7f187800
6975 Node: ChangeLog
\7f189271
6976 Node: Authors
\7f226820
6977 Node: Copying
\7f229095
6978 Node: GNU Free Documentation License
\7f230163
6979 Node: GNU Lesser General Public License
\7f252602